Acharya, P, Nguyen, KD, La, HM, Liu, D & Chen, I-M 2020, 'Nonprehensile Manipulation: a Trajectory-Planning Perspective', IEEE/ASME Transactions on Mechatronics, vol. PP, no. 99, pp. 1-1.
View/Download from: Publisher's site
View description>>
IEEE This paper discusses nonprehensile manipulation of an asymmetric object using a robotic manipulator from a motion planning point of view. Four different aspects of the problem will be analyzed: object stability, motion planning, manipulator control, and experimental validation. Specifically, via an analysis of marginal stability of an object resting on a moving tray, the work establishes the critical accelerations of the manipulator's end-effector, below which the object's stability is guaranteed. These critical accelerations guide the design of the end-effector's motion for successful nonprehensile manipulation of the object. In particular, we propose two methods to formulate polynomial asymmetric s-curve trajectories such that the end- effector completes its motion in minimum time. In one method, the trajectory is divided into segments whose time intervals are then computed via a recursive algorithm. In the other method, we formulate an optimization problem and design the minimum-time trajectory by balancing the trade-off between the travel time and actuator effort. A series of experiments with a robotic arm is designed to validate and compare these motion planning methods in the context of nonprehensile manipulation. In addition, the experimental results demonstrate the advantages of the asymmetric s-curve motion profiles over the traditional symmetric s-curves.
Ayanian, N, Robuffo Giordano, P, Fitch, R, Franchi, A & Sabattini, L 2020, 'Guest editorial: special issue on multi-robot and multi-agent systems', Autonomous Robots, vol. 44, no. 3-4, pp. 297-298.
View/Download from: Publisher's site
Azzam, R, Taha, T, Huang, S & Zweiri, Y 2020, 'Feature-based visual simultaneous localization and mapping: a survey', SN Applied Sciences, vol. 2, no. 2, p. 224.
View/Download from: Publisher's site
Brambley, G & Kim, J 2020, 'Unit dual quaternion‐based pose optimisation for visual runway observations', IET Cyber-Systems and Robotics, vol. 2, no. 4, pp. 181-189.
View/Download from: Publisher's site
Chen, Y, Huang, S & Fitch, R 2020, 'Active SLAM for Mobile Robots With Area Coverage and Obstacle Avoidance', IEEE/ASME Transactions on Mechatronics, vol. 25, no. 3, pp. 1182-1192.
View/Download from: Publisher's site
View description>>
© 1996-2012 IEEE. In this article, we present an active simultaneous localization and mapping (SLAM) framework for a mobile robot to obtain a collision-free trajectory with good performance in SLAM uncertainty reduction and in an area coverage task. Based on a model predictive control framework, these two tasks are solved by the introduction of a control switching mechanism. For SLAM uncertainty reduction, graph topology is used to approximate the original problem as a constrained nonlinear least squares problem. A convex half-space representation is applied to relax nonconvex spatial constraints that represent obstacle avoidance. Using convex relaxation, the problem is solved by a convex optimization method and a rounding procedure based on singular value decomposition. The area coverage task is addressed with a sequential quadratic programming method. A submap joining approach, called linear SLAM, is used to address the associated challenges of avoiding local minima, minimizing control switching, and potentially high computational cost. Finally, various simulations and experiments using an aerial robot are presented that verify the effectiveness of the proposed method, showing that our method produces a more accurate SLAM result and is more computationally efficient compared with multiple existing methods.
Chen, Y, Leighton, B, Zhu, H, Ke, X, Liu, S & Zhao, L 2020, 'Submap-Based Indoor Navigation System for the Fetch Robot', IEEE Access, vol. 8, pp. 81479-81491.
View/Download from: Publisher's site
Chen, Y, Zhao, L, Lee, KMB, Yoo, C, Huang, S & Fitch, R 2020, 'Broadcast Your Weaknesses: Cooperative Active Pose-Graph SLAM for Multiple Robots', IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 2200-2207.
View/Download from: Publisher's site
View description>>
© 2016 IEEE. In this letter, we propose a low-cost, high-efficiency framework for cooperative active pose-graph simultaneous localization and mapping (SLAM) for multiple robots in three-dimensional (3D) environments based on graph topology. Based on the selection of weak connections in pose graphs, this method aims to find the best trajectories for optimal information exchange to repair these weaknesses opportunistically when robots move near them. Based on tree-connectivity, which is greatly related to the D-optimality metric of the Fisher information matrix (FIM), we explore the relationship between measurement (edge) selection and pose-measurement (node-edge) selection, which often occurs in active SLAM, in terms of information increment. The measurement selection problem is formulated as a submodular optimization problem and solved by an exhaustive method using rank-1 updates. We decide which robot takes the selected measurements through a bidding framework where each robot computes its predicted cost. Finally, based on a novel continuous trajectory optimization method, these additional measurements collected by the winning robot are sent to the requesting robot to strengthen its pose graph. In simulations and experiments, we validate our approach by comparing against existing methods. Further, we demonstrate online communication based on offline planning results using two unmanned aerial vehicles (UAVs).
Chen, Y, Zhao, L, Zhang, Y & Huang, S 2020, 'Dense Isometric Non-Rigid Shape-From-Motion Based on Graph Optimization and Edge Selection', IEEE Robotics and Automation Letters, vol. 5, no. 4, pp. 5889-5896.
View/Download from: Publisher's site
View description>>
In this letter, we propose a novel framework for dense isometric non-rigid shape-from-motion (Iso-NRSfM) based on graph topology and edge selection. A weighted undirected graph, of which nodes, edges, and weighted values are respectively the images, the image warps, and the number of the common features, is built. An edge selection algorithm based on maximum spanning tree and sub-modular optimization is presented to pick out the well-connected sub-graph for the warps with multiple images. Using the infinitesimal planarity assumption, the Iso-NRSfM problem is formulated as a graph optimization problem with the virtual measurements, which are based on metric tensor and Christoffel Symbol, and the variables related to the derivatives of the constructed points along the surface. The solution of this graph optimization problem directly leads to the normal field of the shape. Then, using a separable iterative optimization method, we obtain the dense point cloud with texture corresponding to the deformable shape robustly. In the experiments, the proposed method outperforms existing work in terms of constructed accuracy, especially when there exists missing/appearing (changing) data, noisy data, and outliers.
Dai, P, Lu, W, Le, K & Liu, D 2020, 'Sliding Mode Impedance Control for contact intervention of an I-AUV: Simulation and experimental validation', Ocean Engineering, vol. 196, pp. 106855-106855.
View/Download from: Publisher's site
View description>>
© 2019 Elsevier Ltd Position/force control of an Intervention Autonomous Underwater Vehicle (I-AUV) is essential to many underwater intervention tasks (e.g., marine equipment maintenance, underwater welding and so on), and is challenging due to unknown fluid disturbances and model uncertainties. This paper applies the Sliding Mode Impedance Control (SMIC) to the full contact intervention of an I-AUV, from non-contact phase to contact phase. Both computational simulations and practical experiments are conducted to investigate the performance of SMIC. In simulations, considering model uncertainties and detailed fluid disturbances, accurate position and force tracking can be achieved, with the position tracking errors within ±3 × 10−3m and a Root Mean Square Error (RMSE) of 4 × 10−2N in tracking the desired contact force of 10N. For the purpose of experimental validation, the SMIC is implemented on an I-AUV developed in the University of Technology Sydney (UTS). The experimental results demonstrate the SMIC's good performance in the contact intervention of the I-AUV, with the position tracking errors within ±1.6×10−2 m and a RMSE of 0.97N in maintaining the desired contact force of 10N.
Dai, P, Lu, W, Le, K & Liu, D 2020, 'Sliding Mode Impedance Control for contact intervention of an I-AUV: Simulation and experimental validation', Ocean Engineering, vol. 196.
Ding, X, Wang, Y, Xiong, R, Li, D, Tang, L, Yin, H & Zhao, L 2020, 'Persistent Stereo Visual Localization on Cross-Modal Invariant Map', IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 11, pp. 4646-4658.
View/Download from: Publisher's site
View description>>
Autonomous mobile vehicles are expected to perform persistent and accurate localization with low-cost equipment. To achieve this goal, we propose a stereo camera based visual localization method using a modified laser map, which takes the advantage of both the low cost of camera, and high geometric precision of laser data to achieve long-term performance. Considering that LiDAR and camera give measurements of the same environment in different modalities, the cross-modal invariance is investigated to modify the laser map for visual localization. Specifically, a map learning algorithm is introduced to sample the robust subsets in laser maps that are useful for visual localization using multi-session visual and laser data. Further, a generative map model is derived to describe this cross-modal invariance, based on which two types of measurements are defined to model the laser map points as appropriate visual observations. Tightly coupling these measurements within the local bundle adjustment during online sliding-window based visual odometry, the vehicle can achieve robust localization even one year after the map was built. The effectiveness of the proposed method is evaluated on both the public KITTI datasets and self-collected datasets in our campus, which include seasonal, illumination and object variations. On all experimental localization sessions, our method provides satisfactory results, even when the direction is opposite to that in the mapping session, verifying the superior performance of the laser map based visual localization method.
Graham, P, Nikolova, N & Sankaran, S 2020, 'Tension between Leadership Archetypes: Systematic Review to Inform Construction Research and Practice', Journal of Management in Engineering, vol. 36, no. 1, pp. 03119002-03119002.
View/Download from: Publisher's site
View description>>
© 2019 American Society of Civil Engineers. In the literature on construction projects, the role of project managers in maintaining control over tasks and activities has been theorized comprehensively, placing a firm focus on vertical forms of leadership. Increasingly, construction firms are challenged with unprecedented operational uncertainty brought about by changes to project environments, technology, and labor. Similar challenges in other contexts have led to growing research on shared or horizontal approaches to leadership, which have been particularly effective in making organizations more agile in uncertain environments. Through a systematic review of 290 peer-reviewed articles on leadership in construction, this paper considers the extent to which traditional vertical approaches to leadership are supplemented with horizontal and emerging balanced approaches to leadership across six bodies of construction leadership research. It contends that despite evidence for the increasing implementation of horizontal leadership practices on construction projects, vertical leadership theory dominates construction leadership research. In comparison, there is a dearth of research addressing horizontal leadership and scarce consideration of balanced leadership. Based on the review, stronger integration of the balanced leadership archetype in research on leadership in construction is proposed as a logical means of advancing leadership theory in relation to six research vectors.
Hassan, M & Liu, D 2020, 'PPCPP: A Predator–Prey-Based Approach to Adaptive Coverage Path Planning', IEEE Transactions on Robotics, vol. 36, no. 1, pp. 284-301.
View/Download from: Publisher's site
View description>>
© 2004-2012 IEEE. Most of the existing coverage path planning (CPP) algorithms do not have the capability of enabling a robot to handle unexpected changes in the coverage area of interest. Examples of unexpected changes include the sudden introduction of stationary or dynamic obstacles in the environment and change in the reachable area for coverage (e.g., due to imperfect base localization by an industrial robot). Thus, a novel adaptive CPP approach is developed that is efficient to respond to changes in real-time while aiming to achieve complete coverage with minimal cost. As part of the approach, a total reward function that incorporates three rewards is designed where the first reward is inspired by the predator-prey relation, the second reward is related to continuing motion in a straight direction, and the third reward is related to covering the boundary. The total reward function acts as a heuristic to guide the robot at each step. For a given map of an environment, model parameters are first tuned offline to minimize the path length while assuming no obstacles. It is shown that applying these learned parameters during real-time adaptive planning in the presence of obstacles will still result in a coverage path with a length close to the optimized path length. Many case studies with various scenarios are presented to validate the approach and to perform numerous comparisons.
Kong, FH & Manchester, IR 2020, 'Contraction analysis of nonlinear noncausal iterative learning control', Systems & Control Letters, vol. 136, pp. 104599-104599.
View/Download from: Publisher's site
View description>>
Iterative learning control (ILC) is a method for learning input signals for repetitive control tasks. In this paper, we provide a new method based on convex optimization for certifying convergence and estimating convergence rate in ILC schemes involving a nonlinear plant and a noncausal update law, which are common in practice. Using sum-of-squares (SOS) optimization, we compute the convergence rate of an example nonlinear, noncausal ILC system and verify its accuracy in experiment.
Kong, FH, Zhao, J, Zhao, L & Huang, S 2020, 'Analysis of Minima for Geodesic and Chordal Cost for a Minimal 2-D Pose-Graph SLAM Problem', IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 323-330.
View/Download from: Publisher's site
View description>>
© 2016 IEEE. In this letter, we show that for a minimal 2D pose-graph SLAM problem, even in the ideal case of perfect measurements and spherical covariance, using geodesic distance (in 2D, the 'wrap function') to compare angles results in multiple suboptimal local minima. We numerically estimate regions of attraction to these local minima for some examples, give evidence to show that they are of nonzero measure, and that these regions grow in size as noise is added. In contrast, under the same assumptions, we show that the chordal distance representation of angle error has a unique minimum up to periodicity. For chordal cost, we find that initial conditions failing to converge to the global minimum are far fewer, fail because of numerical issues, and do not seem to grow with noise in our examples.
Le Gentil, C, Vidal-Calleja, T & Huang, S 2020, 'Gaussian Process Preintegration for Inertial-Aided State Estimation', IEEE Robotics and Automation Letters, vol. 5, no. 2, pp. 2108-2114.
View/Download from: Publisher's site
View description>>
© 2020 IEEE. In this letter, we present Gaussian Process Preintegration, a preintegration theory based on continuous representations of inertial measurements. A novel use of linear operators on Gaussian Process kernels is employed to generate the proposed Gaussian Preintegrated Measurements (GPMs). This formulation allows the analytical integration of inertial signals on any time interval. Consequently, GPMs are especially suited for asynchronous inertial-aided estimation frameworks. Unlike discrete preintegration approaches, the proposed method does not rely on any explicit motion-model and does not suffer from numerical integration noise. Additionally, we provide the analytical derivation of the Jacobians involved in the first-order expansion for postintegration bias and inter-sensor time-shift correction. We benchmarked the proposed method against existing preintegration methods on simulated data. Our experiments show that GPMs produce the most accurate results and their computation time allows close-to-real-time operations. We validated the suitability of GPMs for inertial-aided estimation by integrating them into a lidar-inertial localisation and mapping framework.
Lu, W & Liu, D 2020, 'A2: Extracting cyclic switchings from DOB-nets for rejecting excessive disturbances', Neurocomputing, vol. 400, pp. 161-172.
View/Download from: Publisher's site
View description>>
© 2020 Reinforcement Learning (RL) is limited in practice by its poor explainability, which is responsible for insufficient trustiness from users, unsatisfied interpretation for human intervention, inadequate analysis for future improvement, etc. This paper seeks to partially characterize the interplay between dynamical environments and a previously-proposed Disturbance OBserver net (DOB-net). The DOB-net is trained via RL and offers optimal control for a set of Partially Observable Markovian Decision Processes (POMDPs). The transition function of each POMDP is largely determined by the environments (excessive external disturbances). This paper proposes an Attention-based Abstraction (A2) approach to extract a finite-state automaton, referred to as a Key Moore Machine Network (KMMN), to capture the switching mechanisms exhibited by the DOB-net in dealing with multiple such POMDPs. A2 first quantizes the controlled platform by learning continuous-discrete interfaces. Then it extracts the KMMN by finding the key hidden states and transitions that attract sufficient attention from the DOB-net. Within the resultant KMMN, three patterns of cyclic switchings (between key hidden states) are found, and saturated controls are shown synchronized with unknown disturbances. Interestingly, the found switchings have previously appeared in the control design for often-saturated systems. They are interpreted via an analogy to the discrete-event subsystem of hybrid control.
Ma, H, Wang, Y, Xiong, R, Kodagoda, S & Tang, L 2020, 'DeepGoal: Learning to drive with driving intention from human control demonstration', Robotics and Autonomous Systems, vol. 127, pp. 103477-103477.
View/Download from: Publisher's site
View description>>
© 2020 Elsevier B.V. Recent research on automotive driving has developed an efficient end-to-end learning mode that directly maps visual input to control commands. However, it models distinct driving variations in a single network, which increases learning complexity and is less adaptive for modular integration. In this paper, we re-investigate human's driving style and propose to learn an intermediate driving intention region to relax the difficulties in end-to-end approach. The intention region follows both road structure in image and direction towards goal in public route planner, which addresses visual variations only and figures out where to go without conventional precise localization. Then the learned visual intention is projected on vehicle local coordinate and fused with reliable obstacle perception to render a navigation score map that is widely used for motion planning. The core of the proposed system is a weakly-supervised cGAN-LSTM model trained to learn driving intention from human demonstration. The adversarial loss learns from limited demonstration data with one local planned route and enables reasoning of multi-modal behaviors with diverse routes while testing. Comprehensive experiments are conducted with real-world datasets. Results indicate the proposed paradigm can produce more consistent motion commands with human demonstration and shows better reliability and robustness to environment change. Our code is available at https://github.com/HuifangZJU/visual-navigation.
Munasinghe, N & Paul, G 2020, 'Ultrasonic-Based Sensor Fusion Approach to Measure Flow Rate in Partially Filled Pipes', IEEE Sensors Journal, vol. 20, no. 11, pp. 6083-6090.
View/Download from: Publisher's site
View description>>
Flow rate measurement in pipes is essential for many applications. Thus, there have been a variety of flow meters developed that incorporate different technologies. However, a typical limitation in flow meters is that the pipe must be full in order to get an accurate flow reading. In many cases, this is not possible for practical reasons. When the pipe is full, ultrasonic flow meters can calculate the flow rate using known properties of the pipe and fluid, namely the cross-section, propagation path and fluid sound velocity. However, when the pipe is only partially filled, the propagation path is unknown which leads to an inability to calculate the correct flow rate. This paper presents a cost-effective sensor fusion approach to extend the capabilities of transit time ultrasonic flow meters to handle such scenarios. The approach determines the propagation path using capacitance-based level sensing, combined with fluid velocities ascertained via an ultrasonic sensor, leading to a significantly more accurate estimation of flow rates. Experiments in low flow rate situations validated the efficacy of the proposed model, with a 92% reduction of mean error in the lowest water height when compared to a conventional ultrasonic flow meter.
Nguyen, L & Miro, JV 2020, 'An Efficient 3-D Model for Remaining Wall Thicknesses of Cast Iron Pipes in Nondestructive Testing', IEEE Sensors Letters, vol. 4, no. 7, pp. 1-4.
View/Download from: Publisher's site
View description>>
© 2017 IEEE. Over 50% of global pipes have been made of cast iron, and most of them are aging. In order to effectively estimate possibilities of their failures, which is paramount for efficiently managing asset infrastructures, it requires the remaining wall thickness (RWT) of a pipe to be known. In fact, RWT of the cast iron pipes can be primarily measured by the magnetism based nondestructive testing technologies though they are quite slow. To speed up the inspection process, it is proposed to sense RWT of a part of a pipe and then employ a model to predict RWT in the rest. Thus, this letter introduces a 3-D model to efficiently represent RWT of a pipe given measurements collected intermittently on the pipe's surface. The proposed model first transforms 3-D cylindrical coordinates to 3-D Cartesian coordinates before modeling RWT by Gaussian processes (GP). The transformation allows GP to work properly on RWT data gathered on a cylindrical pipe and effectively predict RWT at unmeasured locations. Moreover, periodicity of RWT along circumference of the pipe is naturally integrated. The effectiveness of the proposed approach is demonstrated by implementation in two real life inservice aging cast iron pipes, where the obtained results are highly promising.
Nguyen, L & Miro, JV 2020, 'An Efficient 3-D Model for Remaining Wall Thicknesses of Cast Iron Pipes in Nondestructive Testing', IEEE Sensors Letters, vol. 4, no. 7.
View/Download from: Publisher's site
View description>>
© 2017 IEEE. Over 50% of global pipes have been made of cast iron, and most of them are aging. In order to effectively estimate possibilities of their failures, which is paramount for efficiently managing asset infrastructures, it requires the remaining wall thickness (RWT) of a pipe to be known. In fact, RWT of the cast iron pipes can be primarily measured by the magnetism based nondestructive testing technologies though they are quite slow. To speed up the inspection process, it is proposed to sense RWT of a part of a pipe and then employ a model to predict RWT in the rest. Thus, this letter introduces a 3-D model to efficiently represent RWT of a pipe given measurements collected intermittently on the pipe's surface. The proposed model first transforms 3-D cylindrical coordinates to 3-D Cartesian coordinates before modeling RWT by Gaussian processes (GP). The transformation allows GP to work properly on RWT data gathered on a cylindrical pipe and effectively predict RWT at unmeasured locations. Moreover, periodicity of RWT along circumference of the pipe is naturally integrated. The effectiveness of the proposed approach is demonstrated by implementation in two real life inservice aging cast iron pipes, where the obtained results are highly promising.
Nguyen, L & Miro, JV 2020, 'Efficient Evaluation of Remaining Wall Thickness in Corroded Water Pipes Using Pulsed Eddy Current Data', IEEE Sensors Journal, vol. 20, no. 23, pp. 14465-14473.
View/Download from: Publisher's site
View description>>
© 2001-2012 IEEE. In order to analyse failures of an ageing water pipe, some methods such as the loss-of-section require remaining wall thickness (RWT) along the pipe to be fully known, which can be measured by the magnetism based non-destructive evaluation sensors though they are practically slow due to the magnetic penetrating process. That is, fully measuring RWT at every location in a water pipe is not really practical if RWT inspection causes disruption of water supply to customers. Thus, this paper proposes a new data prediction approach that can increase amount of RWT data of a corroded water pipe collected in a given period of time by only measuring RWT on a part (e.g. 20%) of the total pipe surface area and then employing the measurements to predict RWT at unmeasured area. It is proposed to utilize a marginal distribution to convert the non-Gaussian RWT measurements to the standard normally distributed data, which can then be input into a 3-dimensional Gaussian process model for efficiently predicting RWT at unmeasured locations on the pipe. The proposed approach was implemented in two real-life in-service pipes, and the obtained results demonstrate its practicality.
Nikoloska, R, Bykerk, L, Vitanage, D, Valls Miro, J, Chen, F, Wang, Y & Liang, B 2020, 'Enhancing Sydney Water’s leak prevention through acoustic monitoring', Water e-Journal, vol. 5, no. 2, pp. 1-15.
View/Download from: Publisher's site
Ninan, J, Mahalingam, A, Clegg, S & Sankaran, S 2020, 'ICT for external stakeholder management: sociomateriality from a power perspective', Construction Management and Economics, vol. 38, no. 9, pp. 840-855.
View/Download from: Publisher's site
View description>>
External stakeholder support is critical to the success of megaprojects, necessitating strategic engagement, often using Information and Communications Technology (ICT). We conducted 30 semi-structured interviews with a megaproject team and analysed their social media communications with the project community. The findings show three ICT practices used for managing external stakeholders: visualisation, simulation and social mediatisation. Taking a sociomateriality lens we demonstrate how these practices are used for diverse unintended uses to manage external stakeholders. Anchored in a dimensions of power framework, we discuss how these ICT practices were strategically used for persuading, framing and hegemonizing external stakeholders in megaprojects. Theoretically, we highlight the role of ICT for managing external stakeholders over the current use of improving the competitive advantage of internal stakeholders. Practically, social media is used to articulate practices in all the strategic roles, positioning it in a role as a critical ICT tool for external stakeholder management in infrastructure megaprojects.
Otte, M, Sofge, D & Fitch, R 2020, 'Guest editorial: Special issue on robot communication challenges: real-world problems, systems, and methods', Autonomous Robots, vol. 44, no. 1, pp. 1-2.
View/Download from: Publisher's site
Popović, M, Vidal-Calleja, T, Hitz, G, Chung, JJ, Sa, I, Siegwart, R & Nieto, J 2020, 'An informative path planning framework for UAV-based terrain monitoring', Autonomous Robots, vol. 44, no. 6, pp. 889-911.
View/Download from: Publisher's site
View description>>
AbstractUnmanned aerial vehicles represent a new frontier in a wide range of monitoring and research applications. To fully leverage their potential, a key challenge is planning missions for efficient data acquisition in complex environments. To address this issue, this article introduces a general informative path planning framework for monitoring scenarios using an aerial robot, focusing on problems in which the value of sensor information is unevenly distributed in a target area and unknown a priori. The approach is capable of learning and focusing on regions of interest via adaptation to map either discrete or continuous variables on the terrain using variable-resolution data received from probabilistic sensors. During a mission, the terrain maps built online are used to plan information-rich trajectories in continuous 3-D space by optimizing initial solutions obtained by a coarse grid search. Extensive simulations show that our approach is more efficient than existing methods. We also demonstrate its real-time application on a photorealistic mapping scenario using a publicly available dataset and a proof of concept for an agricultural monitoring task.
Qian, K, Liu, H, Valls Miro, J, Jing, X & Zhou, B 2020, 'Hierarchical and parameterized learning of pick-and-place manipulation from under-specified human demonstrations', Advanced Robotics, vol. 34, no. 13, pp. 858-872.
View/Download from: Publisher's site
View description>>
© 2020, © 2020 Informa UK Limited, trading as Taylor & Francis Group and The Robotics Society of Japan. Imitating manipulation skills through observing human demonstrations in everyday life is promising in allowing service robots to be programed quickly, as well as to perform human-like behaviors. Such a Learning by demonstration (LbD) problem is challenging because robots are expected to adapt their learned behaviors to the changes of task parameters and the environment, rather than simply cloning the human teacher's motion. In this paper, we propose a hierarchical and parameterized LbD framework that combines symbolic and trajectory learning of pick-and-place manipulation tasks. We have extended the two-step parameterized learning method with error compensation for learning Environment-adaptive Action Primitives (EaAPs), which is capable of adapting robot's reproduced trajectories to new task instances as well as environmental changes. To arrive at refined plans in situations of under-specified human demonstrations, we propose to model the semantics of demonstrated activities with PDDL-based skill scripts. Therefore, latent motion primitives that are impossible to be learned directly from observing human demonstration in noisy video data can be inferred. The proposed method is implemented as a hierarchical LbD framework and has been evaluated on real robot hardware to illustrate the effectiveness of the proposed approach.
Reid, W, Fitch, R, Göktoğan, AH & Sukkarieh, S 2020, 'Sampling‐based hierarchical motion planning for a reconfigurable wheel‐on‐leg planetary analogue exploration rover', Journal of Field Robotics, vol. 37, no. 5, pp. 786-811.
View/Download from: Publisher's site
View description>>
AbstractReconfigurable mobile planetary rovers are versatile platforms that may safely traverse cluttered environments by morphing their physical geometry. Planning paths for these adaptive robots is challenging due to their many degrees of freedom, and the need to consider potentially continuous platform reconfiguration along the length of the path. We propose a novel hierarchical structure for asymptotically optimal (AO) sampling‐based planners and specifically apply it to the state‐of‐the‐art Fast Marching Tree (FMT*) AO planner. Our algorithm assumes a decomposition of the full configuration space into multiple subspaces, and begins by rapidly finding a set of paths through one such subspace. This set of solutions is used to generate a biased sampling distribution, which is then explored to find a solution in the full configuration space. This technique provides a novel way to incorporate prior knowledge of subspaces to efficiently bias search within existing AO sampling‐based planners. Importantly, probabilistic completeness and asymptotic optimality are preserved. Experimental results in simulation are provided that benchmark the algorithm against state‐of‐the‐art sampling‐based planners without the hierarchical variation. Additional experimental results performed with a physical wheel‐on‐leg platform demonstrate application to planetary rover mobility and showcase how constraints such as actuator failures and sensor pointing may be easily incorporated into the planning problem. In minimizing an energy objective that combines an approximation of the mechanical work required for platform locomotion with that required for reconfiguration, the planner produces intuitive behaviors where the robot dynamically adjusts its footprint, varies its height, and clambers over obstacles using legged locomotion. These results illustrate the generality of the planner in exploiting the platform's mechanical ability to fluidly trans...
Romeijn, T, Wells, B, Wei, D & Paul, G 2020, 'Investigation into the shear property of thin-walled additively manufactured structures using staggered fused filament fabrication', Additive Manufacturing, vol. 35, pp. 101259-101259.
View/Download from: Publisher's site
View description>>
© 2020 Additive manufacturing techniques, such as Fused Filament Fabrication (FFF), are rapidly revolutionising the manufacturing and mining sectors. This paper presents an investigation into the shear behaviour of thin-walled FFF structures, printed via a proposed ‘offset method’. Firstly, an alternative method of filament positioning in material extrusion is proposed, referred to as the ‘offset method’, which aims to reduce the volume of empty cavities between deposited material. Then the shear properties, density properties, and cross-sectional void surface area are compared to structures printed using the aligned printing method. Experimental results on solid printed (no infill) samples, through four different-sized nozzles, have shown the newly proposed method produces a 6.5 % increase in density and a 7.2 % improvement in maximum in-plane shear stress per millimetre increase in nozzle size, compared with the aligned method of FFF. The offset method was found to produce a material with increased interlayer contact, compared to the aligned method, which results in a higher fictitious shear stress modulus. The effect of the increased interlayer contact on the fictitious shear modulus and real shear stress was investigated using a FEM analysis of the unit cells. In short, using the same feedstock material, the offset method produces a stiffer material with a higher fictitious shear strength than the aligned method of FFF printing.
Sankaran, S, Müller, R & Drouin, N 2020, 'Creating a ‘sustainability sublime’ to enable megaprojects to meet the United Nations sustainable development goals', Systems Research and Behavioral Science, vol. 37, no. 5, pp. 813-826.
View/Download from: Publisher's site
View description>>
AbstractDespite cost and schedule overruns and benefits shortfalls, megaprojects (which are large‐scale projects that typically cost over a billion dollars and take years to develop and build) continue to be promoted and built creating a megaproject paradox. Prominent megaproject scholar Bent Flyvbjerg (2014) argued that this could be motivated by four ‘sublimes’—technological, political, economic and aesthetic that drive new megaprojects being put forward despite their poor performance. Recent evidence shows that better governance practices are helping to improve the overall performance of megaprojects. Despite the United Nations setting 17 sustainable development goals (SDGs) to be achieved by 2030, there are severe shortfalls in initiatives from governments, public organizations and private businesses endangering the achievement of targets set for these goals. In addition, time is running out to achieve these goals with only a decade left. The current initiatives contributing to these goals appear to be focused on individual SDGs even though many of these are interrelated. This article proposes that if politicians, engineers and scientists, businesses leaders and design thinkers could be motivated by a ‘sustainability sublime’ to create megaprojects that contribute to SDGs, it could benefit both the society and the planet. It also argues that a more integrated view of UN SDGs and a suitable governance structure should be applied to ensure that megaprojects created as a result of the sustainability sublime deliver benefits towards achieving UN SDGs.
Sankaran, S, Müller, R & Drouin, N 2020, 'Investigating collaboration in project management research: using action research as a meta-methodology', International Journal of Managing Projects in Business, vol. 14, no. 1, pp. 205-230.
View/Download from: Publisher's site
View description>>
PurposeThe purpose of this article is to investigate collaboration in project management research. Although the literature shows an increase in collaboration between scientists and social scientists for various reasons, it is unclear how and why such collaboration takes place in project management research. The literature does show that co-authorship of articles published in project management journals is on the rise due to increased collaboration between researchers in developed countries and emerging economies as well as developing countries. However, no detailed study has been conducted to investigate how such collaboration occurs in practice in project management research. This article addresses this gap.Design/methodology/approachWe use a multi-method approach (action research as a meta-methodology and surveys) using qualitative data to reflect on a successful collaborative externally funded research project. At the end of the study, a survey was used to investigate how collaboration occurred among the 26 researchers involved, who were spread over nine countries to collect data on a sponsored research project led by the authors who were the principal investigators. We also compare our findings from the original project with findings from a second survey of a purposeful sample of ten project management researchers who have conducted or are conducting collaborative research in order to validate our findings.FindingsThrough this study, we were able to compare the reasons for increased collaboration in scientific research reported in the literature with what we learnt from our own experience in collaborating on a large-scale project across geographical boundaries and cultures around ...
Shakor, P, Nejadi, S & Paul, G 2020, 'Investigation into the effect of delays between printed layers on the mechanical strength of inkjet 3DP mortar', Manufacturing Letters, vol. 23, pp. 19-22.
View/Download from: Publisher's site
View description>>
© 2019 Currently, additive manufacturing have enabled to fabricate the three-dimensional models. 3D-Printing technique is a multipurpose process for producing structural members using a sequential layering approach. The “feature quality” of 3DP specimens can be improved by optimising the build constraints. In this paper, a mortar mix powder-base has been prepared that consists of cementitious materials. Experiments are conducted to investigate the effects of different delays in printing time on the mechanical properties of the scaffolds. It has been shown that the compressive stress and strength of printed specimens with a delay of 200 ms were greater than specimens with other delay values.
Shakor, P, Nejadi, S, Paul, G & Sanjayan, J 2020, 'Dimensional accuracy, flowability, wettability, and porosity in inkjet 3DP for gypsum and cement mortar materials', Automation in Construction, vol. 110, pp. 102964-102964.
View/Download from: Publisher's site
View description>>
© 2019 Inkjet (powder-based) 3D Printing is a popular and widely used technology, which can be applied to print a wide range of specimens using different powder materials. This paper discusses the use of inkjet 3DP technology for construction applications using custom-made powder instead of commercial gypsum powder (ZP 151). The paper aims to address the differences between ZP 151 and CP (a custom-made construction-specific cement mortar powder) with regard to powder flowability, wettability, powder bed porosity and apparent porosity in 3DP specimens. An inkjet 3D printer is employed and experimental results verify that ZP 151 has a lower angle of repose, a higher contact angle and noticeably less porosity in the powder bed compared with the CP powder. Additionally, specimens printed with ZP 151 have a lower apparent porosity compared with CP specimens. The wettability for each of the powders was tested using contact angle goniometer, while the Optronis Cam-Recorder was used at 1000 fps at 800 × 600 pixel resolution images for the powder flowability tests. The bulk density tester was utilised to find the apparent porosity in the printed specimens. The paper also discusses the details of the printing procedure and dimensional accuracy of printed specimens.
Shakor, P, Nejadi, S, Sutjipto, S, Paul, G & Gowripalan, N 2020, 'Effects of deposition velocity in the presence/absence of E6-glass fibre on extrusion-based 3D printed mortar', Additive Manufacturing, vol. 32, pp. 101069-101069.
View/Download from: Publisher's site
View description>>
© 2020 Additive Manufacturing (AM) technologies are widely used in various fields of industry and research. Continual research has enabled AM technologies to be considered as a feasible substitute for certain applications in the construction industry, particularly given the advances in the use of glass fibre reinforced mortar. An investigation of the resulting mechanical properties of various mortar mixes extruded using a robotic arm is presented. The nozzle paths were projected via ‘spline’ interpolation to obtain the desired trajectory and deposition velocity in the reference frame of the manipulator. Along each path, various mortar mixes, with and without chopped glass fibre, were deposited at different velocities. Tests were conducted to determine their mechanical performance when incorporated in printed structures with different layers (1, 2, 4 and 6 layers). The results are compared with those of conventional cast-in-place mortar. In this study, the mixes consist of ordinary Portland cement, fine sand, chopped glass fibres (6 mm) and chemical admixtures, which are used to print prismatic- and cubic-shaped specimens. Mechanical strength tests were performed on the printed specimens to evaluate the behaviour of the materials in the presence and absence of glass fibre. Robot end-effector velocity tests were performed to examine the printability and extrudability of the mortar mixes. Finally, horizontal and vertical line printing tests were used to determine the workability, buildability and uniformity of the mortar mix and to monitor the fibre flow directions in the printed specimens. The results show that printed specimens with glass fibre have enhanced compressive strength compared with specimens without glass fibre.
Su, D, Vidal-Calleja, T & Miro, JV 2020, 'Asynchronous microphone arrays calibration and sound source tracking', Autonomous Robots, vol. 44, no. 2, pp. 183-204.
View/Download from: Publisher's site
View description>>
© 2019, Springer Science+Business Media, LLC, part of Springer Nature. In this paper, we proposed an optimisation method to solve the problem of sound source localisation and calibration of an asynchronous microphone array. This method is based on the graph-based formulation of the simultaneous localisation and mapping problem. In this formulation, a moving sound source is considered to be observed from a static microphone array. Traditional approaches for sound source localisation rely on the well-known geometrical information of the array and synchronous readings of the audio signals. Recent work relaxed these two requirements by estimating the temporal offset between pair of microphones based on the assumption that the clock timing of each microphone is exactly the same. This assumption requires the sound cards to be identically manufactured, which in practice is not possible to achieve. Hereby an approach is proposed to jointly estimate the array geometrical information, time offset and clock difference/drift rate of each microphone together with the location of a moving sound source. In addition, an observability analysis of the system is performed to investigate the most suitable configuration for sound source localisation. Simulation and experimental results are presented, which prove the effectiveness of the proposed methodology.
Thiyagarajan, K, Kodagoda, S, Ranasinghe, R, Vitanage, D & Iori, G 2020, 'Robust Sensor Suite Combined With Predictive Analytics Enabled Anomaly Detection Model for Smart Monitoring of Concrete Sewer Pipe Surface Moisture Conditions', IEEE Sensors Journal, vol. 20, no. 15, pp. 8232-8243.
View/Download from: Publisher's site
Ulapane, N, Thiyagarajan, K, Hunt, D & Valls Miro, J 2020, 'Quantifying the Relative Thickness of Conductive Ferromagnetic Materials Using Detector Coil-Based Pulsed Eddy Current Sensors', Journal of Visualized Experiments, vol. 155, no. 155.
View/Download from: Publisher's site
View description>>
Thickness quantification of conductive ferromagnetic materials by means of non-destructive evaluation (NDE) is a crucial component of structural health monitoring of infrastructure, especially for assessing the condition of large diameter conductive ferromagnetic pipes found in the energy, water, oil, and gas sectors. Pulsed eddy current (PEC) sensing, especially detector coil-based PEC sensor architecture, has established itself over the years as an effective means for serving this purpose. Approaches for designing PEC sensors as well as processing signals have been presented in previous works. In recent years, the use of the decay rate of the detector coil-based time domain PEC signal for the purpose of thickness quantification has been studied. Such works have established that the decay rate-based method holds generality to the detector coil-based sensor architecture, with a degree of immunity to factors such as sensor shape and size, number of coil turns, and excitation current. Moreover, this method has shown its effectiveness in NDE of large pipes made of grey cast iron. Following such literature, the focus of this work is explicitly PEC sensor detector coil voltage decay rate-based conductive ferromagnetic material thickness quantification. However, the challenge faced by this method is the difficulty of calibration, especially when it comes to applications such as in situ pipe condition assessment since measuring electrical and magnetic properties of certain pipe materials or obtaining calibration samples is difficult in practice. Motivated by that challenge, in contrast to estimating actual thickness as done by some previous works, this work presents a protocol for using the decay rate-based method to quantify relative thickness (i.e., thickness of a particular location with respect to a maximum thickness), without the requirement for calibration.
Wang, Y, Sun, H, Huang, S & Song, Y 2020, 'Description of stability for linear time‐invariant systems based on the first curvature', Mathematical Methods in the Applied Sciences, vol. 43, no. 2, pp. 486-511.
View/Download from: Publisher's site
View description>>
This paper focuses on using the first curvature κ(t) of trajectory to describe the stability of linear time‐invariant system. We extend the results for two and three‐dimensional systems (Wang, Sun, Song et al, arXiv:1808.00290) to n‐dimensional systems. We prove that for a system , (a) if there exists a measurable set whose Lebesgue measure is greater than zero, such that or does not exist for any initial value in this set, then the zero solution of the system is stable; (b) if the matrix A is invertible, and there exists a measurable set whose Lebesgue measure is greater than zero, such that for any initial value in this set, then the zero solution of the system is asymptotically stable.
Wu, L, Falque, R, Perez-Puchalt, V, Liu, L, Pietroni, N & Vidal-Calleja, TA 2020, 'Skeleton-Based Conditionally Independent Gaussian Process Implicit Surfaces for Fusion in Sparse to Dense 3D Reconstruction.', IEEE Robotics Autom. Lett., vol. 5, no. 2, pp. 1532-1539.
View/Download from: Publisher's site
View description>>
© 2016 IEEE. 3D object reconstructions obtained from 2D or 3D cameras are typically noisy. Probabilistic algorithms are suitable for information fusion and can deal with noise robustly. Consequently, these algorithms can be useful for accurate surface reconstruction. This paper presents an approach to estimate a probabilistic representation of the implicit surface of 3D objects. One of the contributions of the paper is the pipeline for generating an accurate reconstruction, given a set of sparse points that are close to the surface and a dense noisy point cloud. A novel submapping method following the topology of the object is proposed to generate conditional independent Gaussian Process Implicit Surfaces. This allows inference and fusion mechanisms to be performed in parallel followed by information propagation through the submaps. Large datasets can efficiently be processed by the proposed pipeline producing not only a surface but also the uncertainty information of the reconstruction. We evaluate the performance of our algorithm using simulated and real datasets.
Yang, T, Miro, JV, Lai, Q, Wang, Y & Xiong, R 2020, 'Cellular Decomposition for Nonrepetitive Coverage Task With Minimum Discontinuities', IEEE/ASME Transactions on Mechatronics, vol. 25, no. 4, pp. 1698-1708.
View/Download from: Publisher's site
View description>>
A mechanism to derive nonrepetitive coverage path solutions with a proven minimal number of discontinuities is proposed in this work, with the aim to avoid unnecessary, costly end-effector lift-offs for manipulators. The problem is motivated by the automatic polishing of an object. Due to the nonbijective mapping between the workspace and the joint-space, a continuous coverage path in the workspace may easily be truncated in the joint-space, incurring undesirable end-effector lift-offs. Inversely, there may be multiple configuration choices to cover the same point of a coverage path through the solution of the inverse kinematics. The solution departs from the conventional local optimization of the coverage path shape in task space, or choosing appropriate but possibly disconnected configurations, to instead explicitly explore the least number of discontinuous motions through the analysis of the structure of valid configurations in joint-space. The two novel contributions of this article include proof that the least number of path discontinuities is predicated on the surrounding environment, independent from the choice of the actual coverage path; thus, has a minimum. In addition, an efficient finite cellular decomposition method to optimally divide the workspace into the minimum number of cells, each traversable without discontinuities by any arbitrary coverage path within. Extensive simulation examples and real-world results on a 5 DoF manipulator are presented to prove the validity of the proposed strategy in realistic settings.
Yin, H, Wang, Y, Ding, X, Tang, L, Huang, S & Xiong, R 2020, '3D LiDAR-Based Global Localization Using Siamese Neural Network', IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 4, pp. 1380-1392.
View/Download from: Publisher's site
Yoo, C, Lensgraf, S, Fitch, R, Clemon, LM & Mettu, RR 2020, 'Toward Optimal FDM Toolpath Planning with Monte Carlo Tree Search.', CoRR, vol. abs/2002.01631, pp. 4037-4043.
Yu, H, Lu, W, Han, Y, Liu, D & Zhang, M 2020, 'Heterogeneous Dimensionality Reduction for Efficient Motion Planning in High-Dimensional Spaces', IEEE Access, vol. 8, pp. 42619-42632.
View/Download from: Publisher's site
View description>>
© 2013 IEEE. Increasing the dimensionality of the configuration space quickly makes trajectory planning computationally intractable. This paper presents an efficient motion planning approach that exploits the heterogeneous low-dimensional structures of a given planning problem. These heterogeneous structures are obtained via a Dirichlet process (DP) mixture model and together cover the entire configuration space, resulting in more dimensionality reduction than single-structure approaches from the existing literature. Then, a unified low-dimensional trajectory optimization problem is formulated based on the obtained heterogeneous structures and a proposed transversality condition which is further solved via SQP in our implementation. The positive results demonstrate the feasibility and efficiency of our trajectory planning approach on an autonomous underwater vehicle (AUV) and a high-dimensional intervention autonomous underwater vehicle (I-AUV) in cluttered 3D environments.
Zhao, J, Zhao, L, Huang, S & Wang, Y 2020, '2D Laser SLAM With General Features Represented by Implicit Functions', IEEE Robotics and Automation Letters, vol. 5, no. 3, pp. 4329-4336.
View/Download from: Publisher's site
View description>>
© 2016 IEEE. The main contribution of this letter is the problem formulation and algorithm framework for 2D laser SLAM with general features represented by implicit functions. Since 2D laser data reflect the distances from the robot to the boundary of objects in the environment, it is natural to use the boundary of the general objects/features within the 2D environment to describe the features. Implicit functions can be used to represent almost arbitrary shapes from simple (e.g. circle, ellipse, line) to complex (e.g. a cross-section of a bunny model), thus it is worth studying implicit-expressed feature in 2D laser SLAM. In this letter, we clearly formulate the SLAM problem with implicit functions as features, with rigorously computed observation covariance matrix to be used in the SLAM objective function and propose a solution framework. Furthermore, we use ellipses and lines as examples to compare the proposed SLAM method with the traditional pre-fit method (represent the feature using its parameters and pre-fit the laser scan to get the fitted parameter as virtual observations). Simulation and experimental results show that our proposed method has a better performance compared with the pre-fit method and other methods, demonstrating the potential of this new SLAM formulation and method.
Zuo, X, Ye, W, Yang, Y, Zheng, R, Vidal‐Calleja, T, Huang, G & Liu, Y 2020, 'Multimodal localization: Stereo over LiDAR map', Journal of Field Robotics, vol. 37, no. 6, pp. 1003-1026.
View/Download from: Publisher's site
View description>>
AbstractIn this paper, we present a real‐time high‐precision visual localization system for an autonomous vehicle which employs only low‐cost stereo cameras to localize the vehicle with a priori map built using a more expensive 3D LiDAR sensor. To this end, we construct two different visual maps: a sparse feature visual map for visual odometry (VO) based motion tracking, and a semidense visual map for registration with the prior LiDAR map. To register two point clouds sourced from different modalities (i.e., cameras and LiDAR), we leverage probabilistic weighted normal distributions transformation (ProW‐NDT), by particularly taking into account the uncertainty of source point clouds. The registration results are then fused via pose graph optimization to correct the VO drift. Moreover, surfels extracted from the prior LiDAR map are used to refine the sparse 3D visual features that will further improve VO‐based motion estimation. The proposed system has been tested extensively in both simulated and real‐world experiments, showing that robust, high‐precision, real‐time localization can be achieved.
An, B, Huang, S, Chen, Z, Lu, Z, Lu, W & Zhang, Y 1970, 'A 16bit 1MS/s High-Bit Sampling SAR ADC with Improved Binary-Weighted Capacitive Array', 2020 IEEE 5th International Conference on Integrated Circuits and Microsystems (ICICM), 2020 IEEE 5th International Conference on Integrated Circuits and Microsystems (ICICM), IEEE, pp. 267-271.
View/Download from: Publisher's site
Au, W, Sakaue, T & Liu, D 1970, 'A Model for Optimising the Size of Climbing Robots for Navigating Truss Structures', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, Las Vegas, NV, USA.
View/Download from: Publisher's site
View description>>
Truss structures can be found in many buildings and civil infrastructure such as bridges and towers. But as these architectures age, their maintenance is required to keep them structurally sound. A legged robotic solution capable of climbing these structures for maintenance is sought, but determining the size and shape of such a robot to maximise structure coverage is a challenging task. This paper proposes a model in which the size of a multi-legged robot is optimised for coverage in a truss structure. A detailed representation of a truss structure is presented, which forms the novel framework for constraint modelling. With this framework, the overall truss structure coverage is modelled, given a robot's size and its climbing performance constraints. This is set up as an optimisation problem, such that its solution represents the optimum size of the robot that satisfies all constraints. Three case studies of practical climbing applications are conducted to verify the model. By intuitive analysis of the model's output data, the results show that the model accurately applies these constraints in a variety of truss structures.
Azzam, R, Taha, T, Huang, S & Zweiri, Y 1970, 'A Deep Learning Framework for Robust Semantic SLAM', 2020 Advances in Science and Engineering Technology International Conferences (ASET), 2020 Advances in Science and Engineering Technology International Conferences (ASET), IEEE, Dubai, United Arab Emirates, pp. 1-7.
View/Download from: Publisher's site
View description>>
Semantic simultaneous localization and mapping (SLAM) is susceptible to several sources of noise that hinder the accuracy of its trajectory and map estimates. Such sources include inaccurate landmark pose estimation and sensor limitations. In this paper, a novel deep learning based approach is proposed to improve the accuracy of semantic SLAM by reducing the trajectory estimation error. A deep neural network consisting of various non-linear activation functions is structured and pre-trained by means of an unsupervised greedy layer-wise pre-training technique. The network is then fine-tuned using the adaptive moment estimation (Adam) optimizer. The training datasets were collected using several simulated and realtime experiments and are composed of two parts, the estimated trajectory and the corresponding ground truth. Ground truth trajectories were obtained using a motion capture system in realtime experiments. The effectiveness of the proposed approach was shown through simulated experiments, real-time experiments, and a sequence from the Technical University of Munich (TUM) RGB-D dataset. The performance of the deep neural network (DNN) was tested with different pre-training techniques and the proposed unsupervised greedy layer-wise pre-training technique proved to perform the best across training, validation, and testing datasets in terms of reducing the mean absolute trajectory error (ATE).
Best, G & Hollinger, GA 1970, 'Decentralised Self-Organising Maps for Multi-Robot Information Gathering', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, pp. 4790-4797.
View/Download from: Publisher's site
Best, G, Cliff, OM, Patten, T, Mettu, RR & Fitch, R 1970, 'Decentralised Monte Carlo Tree Search for Active Perception', Workshop on the Algorithmic Foundations of Robotics (WAFR), Workshop on the Algorithmic Foundations of Robotics (WAFR), Springer International Publishing, San Francisco, USA, pp. 864-879.
View/Download from: Publisher's site
View description>>
We propose a decentralised variant of Monte Carlo tree search (MCTS) that is suitable for a variety of tasks in multi-robot active perception. Our algorithm allows each robot to optimise its own individual action space by maintaining a probability distribution over plans in the joint-action space. Robots periodically communicate a compressed form of these search trees, which are used to update the locally-stored joint distributions using an optimisation approach inspired by variational methods. Our method admits any objective function defined over robot actions, assumes intermittent communication, and is anytime. We extend the analysis of the standard MCTS for our algorithm and characterise asymptotic convergence under reasonable assumptions. We evaluate the practical performance of our method for generalised team orienteering and active object recognition using real data, and show that it compares favourably to centralised MCTS even with severely degraded communication. These examples support the relevance of our algorithm for real-world active perception with multi-robot systems.
Brian Lee, KM, Martens, W, Khatkar, J, Fitch, R & Mettu, R 1970, 'Efficient Updates for Data Association with Mixtures of Gaussian Processes', 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020 IEEE International Conference on Robotics and Automation (ICRA), IEEE, pp. 335-341.
View/Download from: Publisher's site
View description>>
© 2020 IEEE. Gaussian processes (GPs) enable a probabilistic approach to important estimation and classification tasks that arise in robotics applications. Meanwhile, most GP-based methods are often prohibitively slow, thereby posing a substantial barrier to practical applications. Existing sparse methods to speed up GPs seek to either make the model more sparse, or find ways to more efficiently manage a large covariance matrix. In this paper, we present an orthogonal approach that memoises (i.e. reuses) previous computations in GP inference. We demonstrate that a substantial speedup can be achieved by incorporating memoisation into applications in which GPs must be updated frequently. Moreover, we derive a novel online update scheme for sparse GPs that can be used in conjunction with our memoisation approach for a synergistic improvement in performance. Across three robotic vision applications, we demonstrate between 40-100% speed-up over the standard method for inference in GP mixtures.
Carmichael, M, Khonasty, R, Wilkinson, S & Schork, T 1970, 'The wallbot: A low-cost robot for green wall inspection', Australasian Conference on Robotics and Automation, ACRA, Australasian Conference on Robotics and Automation, ARAA, Brisbane, Australia, pp. 1-7.
View description>>
The benefits of urban green infrastructure, such as attenuating the urban heat island effect and improving air quality, are widely accepted. Regardless, the uptake of green walls (i.e. vertical gardens) is low due to the high costs relating to maintenance and OH&S. These barriers to adoption may be mitigated by using robotics to inspect and maintain green walls. In this work we present the Wallbot, a robotic system to inspect, monitor and aid in the maintenance of green walls. In its current form the system comprises of affordable off-the-shelf components to keep the system cost low. Preliminary development of the system, results of initial tests and findings are presented. The system offers the chance to reduce OH&S issues and maintenance costs associated with green walls.
Carmichael, MG, Khonasty, R, Aldini, S & Liu, D 1970, 'Human Preferences in Using Damping to Manage Singularities During Physical Human-Robot Collaboration', 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020 IEEE International Conference on Robotics and Automation (ICRA), IEEE, Paris, France, pp. 10184-10190.
View/Download from: Publisher's site
View description>>
When a robot manipulator approaches a kinematic singular configuration, control strategies need to be employed to ensure safe and robust operation. If this manipulator is being controlled by a human through physical human-robot collaboration, the choice of strategy for handling singularities can have a significant effect on the feelings and impressions of the user. To date the preferences of humans during physical human-robot collaboration regarding strategies for managing kinematic singularities have yet to be thoroughly explored.This work presents an empirical study of a damping-based strategy for handling singularities with regard to the preferences of the human operator. Two different parameters, damping rate and damping asymmetry, are tested using a double-blind A/B pairwise comparison testing protocol. Participants included two cohorts made up of the general public (n=51) and people working within a robotic research centre (n=18). In total 105 individual trials were performed. Results indicate a preference for a faster, asymmetric damping behavior that slows motions towards singularities whilst allowing for faster motions away.
Darwish, A, Halkon, B, Oberst, S, Fitch, R & Rothberg, S 1970, 'CORRECTION OF LASER DOPPLER VIBROMETER MEASUREMENTS AFFECTED BY SENSOR HEAD VIBRATION USING TIME DOMAIN TECHNIQUES', Proceedings of the XI International Conference on Structural Dynamics, XI International Conference on Structural Dynamics, EASD, Athens, pp. 4842-4850.
View/Download from: Publisher's site
View description>>
Despite widespread use in a variety of areas, in-field applications of laser Doppler vibrometers (LDVs) are still somewhat limited due to their inherent sensitivity to vibration of the instrument sensor head itself. Earlier work, briefly reviewed herein, has shown it to be possible
to subtract the instrument vibration via a number of means, however, it has been difficult up to now to truly compare the performance of these. This is compounded by the constraint that a frequency domain based approach only holds for stationary vibration signals while, particularly for in-field applications, an approach that is also applicable to transient signals is necessary.
This paper therefore describes the development of a novel time domain post-processing based approach for vibrating LDV measurement correction and compares it with the frequency domain counterpart. Results show that, while both techniques offer significant improvements in the corrected LDV signal when compared to a reference accelerometer measurement, the time domain based correction outperforms the frequency domain based method by a factor of eight
Gentil, CL, Tschopp, F, Alzugaray, I, Vidal-Calleja, T, Siegwart, R & Nieto, J 1970, 'IDOL: A Framework for IMU-DVS Odometry using Lines', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems, IEEE, Las Vegas, NV, USA, pp. 5863-5870.
View/Download from: Publisher's site
View description>>
In this paper, we introduce IDOL, an optimization-based framework for IMU-DVSOdometry using Lines. Event cameras, also called Dynamic Vision Sensors (DVSs),generate highly asynchronous streams of events triggered upon illuminationchanges for each individual pixel. This novel paradigm presents advantages inlow illumination conditions and high-speed motions. Nonetheless, thisunconventional sensing modality brings new challenges to perform scenereconstruction or motion estimation. The proposed method offers to leverage acontinuous-time representation of the inertial readings to associate each eventwith timely accurate inertial data. The method's front-end extracts eventclusters that belong to line segments in the environment whereas the back-endestimates the system's trajectory alongside the lines' 3D position byminimizing point-to-line distances between individual events and the lines'projection in the image space. A novel attraction/repulsion mechanism ispresented to accurately estimate the lines' extremities, avoiding theirexplicit detection in the event data. The proposed method is benchmarkedagainst a state-of-the-art frame-based visual-inertial odometry framework usingpublic datasets. The results show that IDOL performs at the same order ofmagnitude on most datasets and even shows better orientation estimates. Thesefindings can have a great impact on new algorithms for DVS.
Hadgraft, RG, Francis, B, Fitch, R, Halkon, B & Brown, T 1970, 'Renewing mechanical and mechatronics programs using studios', SEFI 47th Annual Conference: Varietas Delectat... Complexity is the New Normality, Proceedings, SEFI Annual Conference, SEFI, Budapest, Hungary, pp. 511-522.
View description>>
In a world of rapid change, engineering programs need to adapt to be relevant. This paper addresses the renewal processes for mechanical and mechatronics engineering programs at a large university of technology. The paper sits within a wider curriculum change movement, including all engineering and IT programs at this university. Several meetings have been held over the last 3 years with both industry panels and with academic staff and students to understand the nature of the problem. Using a design-thinking approach, we have explored: global trends, the nature of engineering work and projects, the capabilities required by engineers, and the kinds of capabilities that graduates need to operate confidently in this new world of work. There is a clear need for graduates to be more operational as they move from study to work. Consequently, a major focus on experiential learning is emerging as the key delivery vehicle for new kinds of graduates including projects, studios, and internships. These forms of learning are supported by ready access to online materials as required. A central thread is personalisation of the student learning experience through learning contracts and portfolios. There has been constant demand for change in engineering education for at least the last 20 years. Making change happen, however, is another matter. We are in the fortunate position at this university to have high level support from the Chancellery and the Dean to move our engineering programs to be more relevant to the future. This paper describes the process for engaging our academics, students and industry supporters in that process and will be of interest to many who are grappling with similar transitions.
Hassan, M, Liu, D & Chen, X 1970, 'Squircular-CPP: A Smooth Coverage Path Planning Algorithm based on Squircular Fitting and Spiral Path', 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), IEEE, Boston, MA, USA, pp. 1075-1081.
View/Download from: Publisher's site
View description>>
Coverage path planning (CPP) is essential for applications such as robotic floor cleaning and high-pressure cleaning of surfaces. Smooth CPP algorithms have several benefits including smoother motion of the robot and the reduction of aggressive accelerations and decelerations resulting from sharp turns. In this paper, a novel smooth CPP algorithm is presented which is named Squircular-CPP. This algorithm proposes a squircular shape, which is an intermediate shape between the circle and the square, to fit a target area. Squircular-CPP can also fit a shape between the ellipse and the rectangle. The shape fitting is simple, fast, and analytical and doesn't require a preselection of the shape (i.e., square, circle, ellipse or rectangle). It enables and complements the creation of a smooth spiral path within the fitted shape. Several case studies are presented to demonstrate the effectiveness of the algorithm and to compare it against the popular boustrophedon-based coverage approach and the Deformable Spiral CPP (DSCPP) algorithm.
Hassan, M, Mustafic, D & Liu, D 1970, 'Dec-PPCPP: A Decentralized Predator–Prey-based Approach to Adaptive Coverage Path Planning Amid Moving Obstacles', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, Las Vegas, NV, USA, pp. 11732-11739.
View/Download from: Publisher's site
Hauge, E, Bui, A, Rajalingam, J, Karunatilake, N, Hunt, D, Vitanage, D, Dissanayake, G & Valls Miro, J 1970, 'Robotic Pipe Scanning: Intelligent Internal Toolkit for Critical Water Mains'', Ozwater’20 Papers, OzWater'20 Australia's International Water Conference and Exhibition, Australian Water Association, Adelaide, pp. 1-7.
View description>>
Sydney Water manages a complex water network that includes 5,000 km of large-diameter critical pipelines. To maintain customer satisfaction and minimise loss of water it is essential to manage leaks, breaks through implementing an effective and efficient preventive maintenance and renewal program. Sydney Water in collaboration with the University of Technology Sydney’s Centre for Autonomous Systems (UTS CAS) has developed two world-leading robotic condition assessment tools. These travel inside a dewatered pipe, providing a full 360° wall thickness scan up to 500m in length. Sydney Water has successfully deployed the tools during main failures. It also expects to apply the technology in planned maintenance inspection interventions.
Heon Lee, JJ, Yoo, C, Anstee, S & Fitch, R 1970, 'Hierarchical Planning in Time-Dependent Flow Fields for Marine Robots', 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020 IEEE International Conference on Robotics and Automation (ICRA), IEEE, Paris, France (Virtual), pp. 885-891.
View/Download from: Publisher's site
View description>>
We present an efficient approach for finding shortest paths in flow fields that vary as a sequence of flow predictions over time. This approach is applicable to motion planning for slow marine robots that are subject to dynamic ocean currents. Although the problem is NP-hard in general form, we incorporate recent results from the theory of finding shortest paths in time-dependent graphs to construct a polynomial-time algorithm that finds continuous trajectories in time-dependent flow fields. The algorithm has a hierarchical structure where a graph is constructed with time-varying edge costs that are derived from sets of continuous trajectories in the underlying flow field. We show that the continuous algorithm retains the time complexity and path quality properties of the discrete graph solution, and demonstrate its application to surface and underwater vehicles including a traversal along the East Australian Current with an autonomous marine vehicle. Results show that the algorithm performs efficiently in practice and can find paths that adapt to changing ocean currents. These results are significant to marine robotics because they allow for efficient use of time-varying ocean predictions for motion planning.
Jayasuriya, M, Arukgoda, J, Ranasinghe, R & Dissanayake, G 1970, 'Localising PMDs through CNN Based Perception of Urban Streets.', ICRA, IEEE International Conference on Robotics and Automation, IEEE, Paris, France, pp. 6454-6460.
View/Download from: Publisher's site
View description>>
The main contribution of this paper is a novel Extended Kalman Filter (EKF) based localisation scheme that fuses two complementary approaches to outdoor vision based localisation. This EKF is aided by a front end consisting of two Convolutional Neural Networks (CNNs) that provide the necessary perceptual information from camera images. The first approach involves a CNN based extraction of information corresponding to artefacts such as curbs, lane markings, and manhole covers to localise on a vector distance transform representation of a binary image of these ground surface boundaries. The second approach involves a CNN based detection of common environmental landmarks such as tree trunks and light poles, which are represented as point features on a sparse map. Utilising CNNs to obtain higher level information about the environment enables this framework to avoid the typical pitfalls of common vision based approaches that use low level hand crafted features for localisation. The EKF framework makes it possible to deal with false positives and missed detections that are inevitable in a practical CNN, to produce a location estimate together with its associated uncertainty. Experiments using a Personal Mobility Device (PMD) driven in typical suburban streets are presented to demonstrate the effectiveness of the proposed localiser.
Jayasuriya, M, Arukgoda, J, Ranasinghe, R & Dissanayake, G 1970, 'Towards Adapting Autonomous Vehicle Technology for the Improvement of Personal Mobility Devices', 2020 5th International Conference on Innovative Technologies in Intelligent Systems and Industrial Applications (CITISIA), 2020 5th International Conference on Innovative Technologies in Intelligent Systems and Industrial Applications (CITISIA), IEEE, Sydney, Australia, pp. 1-7.
View/Download from: Publisher's site
View description>>
Personal Mobility Devices (PMDs) incorporated with autonomy, have great potential in becoming an essential building block of smart transportation infrastructures of the future. However, autonomous vehicle technologies currently employ large and expensive sensors / computers and resource intensive algorithms, which are not suitable for low cost, small form factor PMDs. In this paper, a mobility scooter is retrofitted with a low cost sensing and computing package with the aim of achieving autonomous driving capability. As a first step, a novel, real time, low cost and resource efficient vision only localisation framework based on Convolutional Neural Network (CNN) oriented feature extraction and extended Kalman filter oriented state estimation is presented. Real world experiments in a suburban environment are presented to demonstrate the effectiveness of the proposed localisation framework.
Jayasuriya, M, Ranasinghe, R & Dissanayake, G 1970, 'Active Perception for Outdoor Localisation with an Omnidirectional Camera.', IROS, 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems, IEEE, Las Vegas, NV, USA, pp. 4567-4574.
View/Download from: Publisher's site
View description>>
This paper presents a novel localisation framework based on an omnidirectional camera, targeted at outdoor urban environments. Bearing only information to persistent and easily observable high-level semantic landmarks (such as lamp-posts, street-signs and trees) are perceived using a Convolutional Neural Network (CNN). The framework utilises an information theoretic strategy to decide the best viewpoint to serve as an input to the CNN instead of the full 360° coverage offered by an omnidirectional camera, in order to leverage the advantage of having a higher field of view without compromising on performance. Environmental landmark observations are supplemented with observations to ground surface boundaries corresponding to high-level features such as manhole covers, pavement edges and lane markings extracted from a second CNN. Localisation is carried out in an Extended Kalman Filter (EKF) framework using a sparse 2D map of the environmental landmarks and Vector Distance Transform (VDT) based representation of the ground surface boundaries. This is in contrast to traditional vision only localisation systems that have to carry out Visual Odometry (VO) or Simultaneous Localisation and Mapping (SLAM), since low level features (such as SIFT, SURF, ORB) do not persist over long time frames due to radical appearance changes (illumination, occlusions etc) and dynamic objects. As the proposed framework relies on highlevel persistent semantic features of the environment, it offers an opportunity to carry out localisation on a prebuilt map, which is significantly more resource efficient and robust. Experiments using a Personal Mobility Device (PMD) driven in a representative urban environment are presented to demonstrate and evaluate the effectiveness of the proposed localiser against relevant state of the art techniques.
Jiao, Y, Wang, Y, Fu, B, Tan, Q, Chen, L, Wang, M, Huang, S & Xiong, R 1970, 'Globally optimal consensus maximization for robust visual inertial localization in point and line map', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, pp. 4631-4638.
View/Download from: Publisher's site
Kamal, A & Miro, JV 1970, 'Monocular end-to-end vehicle pose estimation for car manufacturing', Australasian Conference on Robotics and Automation, ACRA, Australasian Conference on Robotics and Automation (ACRA) 2020, Australian Robotics and Automation Association (ARAA), Brisbane, QLD, Australia.
View description>>
An efficient vehicle pose estimation method from a monocular camera to assist paint defect identification in a car manufacturing setting is presented in this paper. Inspired by promising results reported by the self-driving car community with end-to-end schemes, a cascaded deep neural network is proposed for rapid estimation of both translation and rotation of a moving vehicle along a production line, achieving pose estimate average errors below 1.0cm in translation and 0.009◦ in rotation on a ground-truth synthetic database. Notably compelling for the purpose of potential deployment in real factory settings is the ability to infer poses within 1 second. Comprehensive experiments are presented to determine the most accurate camera configuration, and comparisons to traditional two-stage iterative image processing and pose optimisation methods are also provided to demonstrate the network’s superior performance in provided accurate vehicle pose estimates in real-time.
Katuwandeniya, K, Miro, JV & Dantanarayana, L 1970, 'End-to-End Joint Intention Estimation for Shared Control Personal Mobility Navigation', 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), IEEE, Shenzhen, PEOPLES R CHINA, pp. 1-6.
View/Download from: Publisher's site
View description>>
Advancements in technology propose a future where systems work collaboratively sharing the same workspace as humans. Navigation is one such crucial aspect of daily life where collaborative technologies can offer major assistance. Ageing population dictates a likely increase in personal mobility devices (PMDs), whilst autonomous cars are bringing intelligent vehicles to the road today. However, in such scenarios the expected assistance can only be given if the device is aware of its user's intention, so that controls can be applied in a tightly collaborative manner. Moreover, they should be robust to different environments, users and mobile platforms. A user driven navigation framework is proposed in this work to complement end-to-end sensing-only solutions to estimate controls as joint intention from vehicle states and user inputs. The solution is proven to be an improvement over similar strategies that rely on exteroceptive data and omit inputs from the driving agent. Furthermore, the developed framework is proven capable of transferring the learning into different environments and mobility platforms using a small amount of training data. Data from the autonomous driving community (Udacity dataset) and other obtained in-house with an instrumented power wheelchair are given to demonstrate the validity of the proposed approach.
Khosoussi, K, Sukhatme, GS, Huang, S & Dissanayake, G 1970, 'Designing Sparse Reliable Pose-Graph SLAM: A Graph-Theoretic Approach', International Workshop on the Algorithmic Foundations of Robotics, International Workshop on the Algorithmic Foundations of Robotics, pp. 17-32.
View/Download from: Publisher's site
View description>>
In this paper, we aim to design sparse D-optimal (determinantoptimal) pose-graph SLAM problems through the synthesis of sparse graphs with the maximum weighted number of spanning trees. Characterizing graphs with the maximum number of spanning trees is an open problem in general. To tackle this problem, several new theoretical results are established in this paper, including the monotone log-submodularity of the weighted number of spanning trees. By exploiting these structures, we design a complementary pair of near-optimal efficient approximation algorithms with provable guarantees. Our theoretical results are validated using random graphs and a publicly available pose-graph SLAM dataset.
Kim, J, Bhambhani, Y, Byun, H & Johansen, TA 1970, 'Cascaded nonlinear attitude observer and simultaneous localisation and mapping', Australasian Conference on Robotics and Automation, ACRA, Australasian Conference on Robotics and Automation, ARAA, Brisbane, pp. 1-6.
View description>>
This paper presents a novel integration of the nonlinear observer theory and simultaneous localisation and mapping for aerial navigation applications. This extends the previous work by the authors in which a nonlinear observer was applied to the attitude estimation and integrated navigation problem. The key novelty of this work is in the feedback correction mechanism from the linear SLAM estimator to the nonlinear observer, which enables the attitude correction from the feature position measurements. We utilise the relationship between the acceleration error and the attitude error, and the pseudo-inverse of a skew-symmetric matrix for the attitude feedback. Lyapunov-based stability analysis is provided for a simplified model without considering the gyroscope bias. Flight dataset is used to confirm the method. Thanks to the robustness of the nonlinear observer and the optimal linear estimator, the vehicle pose and map features are estimated effectively.
Kim, J, Byun, H, Guivant, J & Johansen, TA 1970, 'Compressed Pseudo-SLAM: Pseudorange integrated generalised compressed SLAM', Australasian Conference on Robotics and Automation, ACRA, Australasian Conference on Robotics and Automation, ARAA, Brisbane.
View description>>
This paper addresses the fusion of the pseudorange/pseudorange rate observations from global navigation satellite system (GNSS), and the inertial-visual simultaneous localisation and mapping (SLAM) to achieve reliable navigation of unmanned aerial vehicles (UAVs). This work extends the previous work on a simulation-based study [Kim et al.(2017)], and evaluates the method to a flight dataset collected from a fixed-wing UAV platform. We propose to use the generalised compressed filter which can effectively accumulate the information gain acquired from a local map, and update the global map in a much lower rate. The fusion filter also models and estimates the receiver clock and drift, which is crucial to integrate the pseudorange and pseudorange rate measurements. Evaluation results will show that the horizontal navigation error is effectively constrained even with 1 satellite vehicle and 1 landmark observations, thanks to the direct fusion of pseudorange and vision data.
Kiss, SH, To, KYC, Yoo, C, Fitch, R & Alempijevic, A 1970, 'Minimally Invasive Social Navigation', Australasian Conference on Robotics and Automation 2019, Australasian Conference on Robotics and Automation, ARAA, Adelaide, Australia, pp. 1-7.
View description>>
Integrating mobile robots into human society involves the fundamental problemof navigation in crowds. This problem has been studied by considering thebehaviour of humans at the level of individuals, but this representation limitsthe computational efficiency of motion planning algorithms. We explore the ideaof representing a crowd as a flow field, and propose a formal definition ofpath quality based on the concept of invasiveness; a robot should attempt tonavigate in a way that is minimally invasive to humans in its environment. Wedevelop an algorithmic framework for path planning based on this definition andpresent experimental results that indicate its effectiveness. These resultsopen new algorithmic questions motivated by the flow field representation ofcrowds and are a necessary step on the path to end-to-end implementations.
Le Gentil, C, Vayugundla, M, Giubilato, R, Sturzl, W, Vidal-Calleja, T & Triebel, R 1970, 'Gaussian Process Gradient Maps for Loop-Closure Detection in Unstructured Planetary Environments', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, Las Vegas, NV, USA, pp. 1895-1902.
View/Download from: Publisher's site
View description>>
The ability to recognize previously mapped locations is an essential feature for autonomous systems. Unstructured planetary-like environments pose a major challenge to these systems due to the similarity of the terrain. As a result, the ambiguity of the visual appearance makes state-of-the-art visual place recognition approaches less effective than in urban or man-made environments. This paper presents a method to solve the loop closure problem using only spatial information. The key idea is to use a novel continuous and probabilistic representations of terrain elevation maps. Given 3D point clouds of the environment, the proposed approach exploits Gaussian Process (GP) regression with linear operators to generate continuous gradient maps of the terrain elevation information. Traditional image registration techniques are then used to search for potential matches. Loop closures are verified by leveraging both the spatial characteristic of the elevation maps (SE (2) registration) and the probabilistic nature of the GP representation. A submap-based localization and mapping framework is used to demonstrate the validity of the proposed approach. The performance of this pipeline is evaluated and benchmarked using real data from a rover that is equipped with a stereo camera and navigates in challenging, unstructured planetary-like environments in Morocco and on Mt. Etna.
Le, DT, Sutjipto, S, Lai, Y & Paul, G 1970, 'Intuitive Virtual Reality based Control of a Real-world Mobile Manipulator', 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), IEEE, Shenzhen, China.
View/Download from: Publisher's site
View description>>
This paper presents an integration of Virtual Reality (VR) interfaces with the control system of a real-world mobile manipulator, ultimately facilitating a natural and intuitive method for human-robot interaction. VR’s ability to track movements in 3D space and translate performed motions provide an intuitive platform for users to explore and interact with the virtual environment. Coupled with intuitive controls, such as grabbing and pointing, the VR platform provides a compelling advantage that can be used to solve limitations of traditional remote robot teleoperation methods.This paper summarises the system implemented, which includes a simulation of the robot in Unity3d, as well as analyses critical results of accuracy and performance, from experiments with users of various experience levels. The method used for measuring accuracy with a simulated robot presented a utilitarian validation for contrasting the difference between 2D and VR 3D interfaces. Users’ performance and experience under various levels of control latency, which is a crucial factor in remote online robot control, were also measured.
Le, K, To, A, Leighton, B, Hassan, M & Liu, D 1970, 'The SPIR: An Autonomous Underwater Robot for Bridge Pile Cleaning and Condition Assessment', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, pp. 1725-1731.
View/Download from: Publisher's site
Lee, C, Best, G & Hollinger, GA 1970, 'Optimal Deployment of Multiple Passenger Robots using Sequential Stochastic Assignment', RSS Workshop on Heterogeneous Multi-Robot Task Allocation and Coordination.
Liu, L, Zhang, T, Liu, Y, Leighton, B, Zhao, L, Huang, S & Dissanayake, G 1970, 'Parallax Bundle Adjustment on Manifold with Improved Global Initialization', Springer Proceedings in Advanced Robotics (SPAR), International Workshop on the Algorithmic Foundations of Robotics, Springer International Publishing, Mérida, México, pp. 621-638.
View/Download from: Publisher's site
View description>>
In this paper we present a novel extension to the parallax feature based bundle adjustment (BA). We take parallax BA into a manifold form (PMBA) along with an observation-ray based objective function. This formulation faithfully mimics the projective nature in a camera’s image formation, resulting in a stable optimization configuration robust to low-parallax features. Hence it allows use of fast Dogleg optimization algorithm, instead of the usual Levenberg Marquardt. This is particularly useful in urban SLAM in which diverse outdoor environments and collinear motion modes are prevalent. Capitalizing on these properties, we propose a global initialization scheme in which PMBA is simplified into a pose-graph problem. We show that near-optimal solution can be achieved under low-noise conditions. With simulation and a series of challenging publicly available real datasets, we demonstrate PMBA’s superior convergence performance in comparison to other BA methods. We also demonstrate, with the “Bundle Adjustment in the Large” datasets, that our global initialization process successfully bootstrap the full BA in mapping many sequential or out-of-order urban scenes.
Maleki, B, Alempijevic, A & Vidal-Calleja, T 1970, 'Continuous Optimization Framework for Depth Sensor Viewpoint Selection', Workshop on the Algorithmic Foundations of Robotics, Workshop on the Algorithmic Foundations of Robotics, Springer International Publishing, Merida, Mexico, pp. 357-372.
View/Download from: Publisher's site
View description>>
Distinguishing differences between areas represented with point cloud data is generally approached by choosing a optimal viewpoint. The most informative view of a scene ultimately enables to have the optimal coverage over distinct points both locally and globally while accounting for the distance to the foci of attention. Measures of surface saliency, related to curvature inconsistency, extenuate differences in shape and are coupled with viewpoint selection approaches. As there is no analytical solution for optimal viewpoint selection, candidate viewpoints are generally discretely sampled and evaluated for information and require (near) exhaustive combinatorial searches. We present a consolidated optimization framework for optimal viewpoint selection with a continuous cost function and analytically derived Jacobian that incorporates view angle, vertex normals and measures of task related surface information relative to viewpoint. We provide a mechanism in the cost function to incorporate sensor attributes such as operating range, field of view and angular resolution. The framework is evaluated as competing favorably with the state-of-the-art approaches to viewpoint selection while significantly reducing the number of viewpoints to be evaluated in the process.
Mehami, J, Vidal-Calleja, T & Alempijevic, A 1970, 'Observability driven Multi-modal Line-scan Camera Calibration', 2020 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI), 2020 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI), IEEE, Karlsruhe, Germany, pp. 285-290.
View/Download from: Publisher's site
View description>>
© 2020 IEEE. Multi-modal sensors such as hyperspectral line-scan and frame cameras can be incorporated into a single camera system, enabling individual sensor limitations to be compensated. Calibration of such systems is crucial to ensure data from one modality can be related to the other. The best known approach is to capture multiple measurements of a known planar pattern, which are then used to optimize calibration parameters through non-linear least squares. The confidence in the optimized parameters is dependent on the measurements, which are contaminated by noise due to sensor hardware. Understanding how this noise transfers through the calibration is essential, especially when dealing with line-scan cameras that rely on measurements to extract feature points. This paper adopts a maximum likelihood estimation method for propagating measurement noise through the calibration, such that the optimized parameters are associated with an estimate of uncertainty. The uncertainty enables development of an active calibration algorithm, which uses observability to selectively choose images that improve parameter estimation. The algorithm is tested in both simulation and hardware, then compared to a naive approach that uses all images to calibrate. The simulation results for the algorithm show a drop of 26.4% in the total normalized error and 46.8% in the covariance trace. Results from the hardware experiments also show a decrease in the covariance trace, demonstrating the importance of selecting good measurements for parameter estimation.
Miro, JV, Munoz, F & Miguel, FI 1970, 'An Arc-Shaped Rotating Magnet Solution for 3D Localisation of a Drug Delivery Capsule Robot', 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), IEEE, USA, pp. 520-527.
View/Download from: Publisher's site
View description>>
A method to estimate the three-dimensional (3D) position of a capsule robot used to deliver drugs in the gastrointestinal tract is proposed in this paper. By exploiting the unique characteristics of the rotating magnetic field created by an array of tangentially magnetised arc-shaped permanent magnets (ASMs), and its analytical formulation, a capsule robot equipped with on-board Hall-effect sensors can measure the rotating magnetic fields created to infer its pose. Extensive validation results provided from a small rotating ASM experimental rig built to test the concept, and a complementary robotic setup for large scale testing are supplied. Given the proven homothetic transformations of magnetic fields, this work demonstrates with validated practical experimentation in a scaled-down rig (1/10), that a full rotation of the ASMs about one axis is sufficient to obtain a mean pose error <10 mm in a magnetic system operating in scaled workspaces up to 250 mm, relevant for clinical use of capsule robots inside human bodies.
Munasinghe, N & Paul, G 1970, 'Integrated 3-D printable temperature sensor for advanced manufacturing', Australasian Conference on Robotics and Automation, ACRA, Australasian Conference on Robotics and Automation, ARAA, Queensland, Australia.
View description>>
As technology continues to develop at a rapid pace, the world progresses towards the fourth industrial revolution, Industry 4.0, with advancements in automation and machine intelligence, as well as manufacturing breakthroughs leading to more efficient and advanced methods. Additive manufacturing (AM), also known as 3D printing, is a type of manufacturing method that has experienced great development and has revolutionised end-product manufacturing. The authors are involved in a project to develop a large-scale industrial 3D printer to print equipment called a Gravity Separation Spiral (GSS), and in an effort to make the equipment “smart”, sensors need to be embedded inside to monitor the operating conditions remotely. This paper presents a temperature sensor able to be printed by a multi-material 3D printer, into 3D printed equipment. In this method, a conductive carbon-based filament has been used to print temperature-sensitive traces inside a Polylactic Acid (PLA) base. The printed sensor was temperature tested in a controlled environment using a programmable heat pad, and the change in resistance has been measured as a voltage change using a data acquisition device. Tests were conducted within in the expected operating range, between 25 ℃ and 36 ℃, and the absolute temperature error was found to be less than ±2 ℃.
Munasinghe, N & Paul, G 1970, 'Path Planning for Robot Based Radial Advanced Manufacturing Using Print Space Sampling', 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), IEEE, Shenzhen, China, pp. 854-859.
View/Download from: Publisher's site
View description>>
The world is embracing the fourth industrial revolution, Industry 4.0, which is enabling businesses to improve efficiency and optimise operations. The authors are part of a team that is researching and developing a large-scale industrial 3D printer to print smart, bespoke equipment called Gravity Separation Spirals (GSS). GSS are used in mining to separate minerals from the slurry. The printer under development employs two industrial robot arms mounted on vertical rails and the print direction is around a vertical rotating column in a radial direction. This paper presents a cost-based path planning method using print-space sampling to optimise distance error and manipulability during a printhead’s radial path as it travels outwards from the central column. Manipulability, distance error and rotation error have been calculated for each sampled point and a weighted cost function has been used to determine the optimal path. Simulated results show that this method reduces the instances of print failure and improves the overall manipulability of the robot during printing.
Nguyen, DDK, Lai, Y, Sutjipto, S & Paul, G 1970, 'Hybrid Multi-Robot System for Drilling and Blasting Automation', 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), IEEE, Shenzhen, China, pp. 79-84.
View/Download from: Publisher's site
View description>>
Multi-robot systems possess the potential of becoming the next generation of robots in the mining industry due to their robustness and scalability. However, they present challenges for the system to efficiently allocate tasks to each robot and allow them to navigate toward their targets safely. This paper introduces a hybrid approach method for a multi-robot system, alongside with a case study in drilling and blasting automation. A Centralized Control Unit delegates tasks and information among the robots in the system, each equipped with a decentralized motion planner that supports cooperative inter- robot collision avoidance. The proposed system inherits the advantage of a centralized multi-robot system in providing a time-wise optimal solution; while also possessing the computational benefit and scalability of a decentralized system. Simulations were conducted to validate the proposed method and discuss insights into the efficacy and performance of the proposed method.
Popovic, M, Vidal-Calleja, T, Chung, JJ, Nieto, J & Siegwart, R 1970, 'Informative Path Planning for Active Field Mapping under Localization Uncertainty', 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020 IEEE International Conference on Robotics and Automation (ICRA), IEEE, Paris, France (Virtual), pp. 10751-10757.
View/Download from: Publisher's site
View description>>
Information gathering algorithms play a key role in unlocking the potential of robots for efficient data collection in a wide range of applications. However, most existing strategies neglect the fundamental problem of the robot pose uncertainty, which is an implicit requirement for creating robust, high-quality maps. To address this issue, we introduce an informative planning framework for active mapping that explicitly accounts for the pose uncertainty in both the mapping and planning tasks. Our strategy exploits a Gaussian Process (GP) model to capture a target environmental field given the uncertainty on its inputs. For planning, we formulate a new utility function that couples the localization and field mapping objectives in GP-based mapping scenarios in a principled way, without relying on manually-tuned parameters. Extensive simulations show that our approach outperforms existing strategies, reducing mean pose uncertainty and map error. We present a proof of concept in an indoor temperature mapping scenario.
Prabowo, YA, Ranasinghe, R, Dissanayake, G, Riyanto, B & Yuliarto, B 1970, 'A Bayesian approach for gas source localization in large indoor environments', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, pp. 4432-4437.
View/Download from: Publisher's site
Reid, W, Fitch, R, Göktoǧgan, AH & Sukkarieh, S 1970, 'Motion Planning for Reconfigurable Mobile Robots Using Hierarchical Fast Marching Trees', Algorithmic Foundations of Robotics XII Proceedings of the Twelfth Workshop on the Algorithmic Foundations of Robotics, Springer International Publishing, WAFR, pp. 656-671.
View/Download from: Publisher's site
View description>>
Reconfigurable mobile robots are versatile platforms that may safely traverse cluttered environments by morphing their physical geometry. However, planning paths for these robots is challenging due to their many degrees of freedom. We propose a novel hierarchical variant of the Fast Marching Tree (FMT*) algorithm. Our algorithm assumes a decomposition of the full state space into multiple sub-spaces, and begins by rapidly finding a set of paths through one such sub-space. This set of solutions is used to generate a biased sampling distribution, which is then explored to find a solution in the full state space. This technique provides a novel way to incorporate prior knowledge of sub-spaces to efficiently bias search within the existing FMT* framework. Importantly, probabilistic completeness and asymptotic optimality are preserved. Experimental results are provided for a reconfigurable wheel-on-leg platform that benchmark the algorithm against state-of-the-art samplingbased planners. In minimizing an energy objective that combines the mechanical work required for platform locomotion with that required for reconfiguration, the planner produces intuitive behaviors where the robot dynamically adjusts its footprint, varies its height, and clambers over obstacles using legged locomotion. These results illustrate the generality of the planner in exploiting the platform’s mechanical ability to fluidly transition between various physical geometric configurations, and wheeled/legged locomotion modes.
Saroya, M, Best, G & Hollinger, GA 1970, 'Online Exploration of Tunnel Networks Leveraging Topological CNN-based World Predictions', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE.
View/Download from: Publisher's site
Scheide, E, Best, G & Hollinger, GA 1970, 'Learning Behavior Trees for Robotic Task Planning by Monte Carlo Search over a Formal Grammar', RSS Workshop on Learning (in) Task and Motion Planning.
Singh, AK, Aldini, S, Leong, D, Wang, Y-K, Carmichael, MG, Liu, D & Lin, C-T 1970, 'Prediction Error Negativity in Physical Human-Robot Collaboration', 2020 8TH INTERNATIONAL WINTER CONFERENCE ON BRAIN-COMPUTER INTERFACE (BCI), 8th International Winter Conference on Brain-Computer Interface (BCI), IEEE, SOUTH KOREA, Tech Univ Berlin, Korea Univ Machine Learning Grp, BK21 Plus Global Leader, Gangwon, pp. 58-63.
View/Download from: Publisher's site
Singh, AK, Aldini, S, Leong, D, Wang, Y-K, Carmichael, MG, Liu, D & Lin, C-T 1970, 'Prediction Error Negativity in Physical Human-Robot Collaboration', 2020 8th International Winter Conference on Brain-Computer Interface (BCI), 2020 8th International Winter Conference on Brain-Computer Interface (BCI), IEEE, Gangwon, Korea (South), pp. 1-6.
View/Download from: Publisher's site
View description>>
Cognitive conflict is a fundamental phenomenon of human cognition, particularly during interaction with the real world. Understanding and detecting cognitive conflict can help to improve interactions in a variety of applications, such as in human-robot collaboration (HRC), which involves continuously guiding the semi-autonomous robot to perform a task in given settings. There have been several works to detect cognitive conflict in HRC but without physical control settings. In this work, we have conducted the first study to explore cognitive conflict using prediction error negativity (PEN) in physical human-robot collaboration (pHRC). Our results show that there was a statistically significant (p =. 047) higher PEN for conflict condition compared to normal conditions, as well as a statistically significant difference between different levels of PEN (p =. 020). These results indicate that cognitive conflict can be detected in pHRC settings and, consequently, provide a window of opportunities to improve the interaction in pHRC.
Song, J, Bai, F, Zhao, L, Huang, S & Xiong, R 1970, 'Efficient two step optimization for large embedded deformation graph based SLAM', 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020 IEEE International Conference on Robotics and Automation (ICRA), IEEE, pp. 9419-9425.
View/Download from: Publisher's site
View description>>
© 2020 IEEE. Embedded deformation graph is a widely used technique in deformable geometry and graphical problems. Although the technique has been transmitted to stereo (or RGB-D) camera based SLAM applications, it remains challenging to compromise the computational cost as the model grows. In practice, the processing time grows rapidly in accordance with the expansion of maps. In this paper, we propose an approach to decouple the nodes of deformation graph in large scale dense deformable SLAM and keep the estimation time to be constant. We observe that only partial deformable nodes in the graph are connected to visible points. Based on this fact, the sparsity of the original Hessian matrix is utilized to split the parameter estimation into two independent steps. With this new technique, we achieve faster parameter estimation with amortized computation complexity reduced from O(n2) to almost O(1). As a result, the computational cost barely increases as the map keeps growing. Based on our strategy, the computational bottleneck in large scale embedded deformation graph based applications will be greatly mitigated. The effectiveness is validated by experiments, featuring large scale deformation scenarios.
Sutjipto, S, Lai, Y, Carmichael, MG & Paul, G 1970, 'Fitts’ law in the presence of interface inertia', 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), 2020 42nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) in conjunction with the 43rd Annual Conference of the Canadian Medical and Biological Engineering Society, IEEE, Montreal, QC, Canada, Canada, pp. 4749-4752.
View/Download from: Publisher's site
View description>>
Collaborative robots are advancing the healthcare frontier, in applications such as rehabilitation and physical therapy. Effective physical collaboration in human-robot systems require an understanding of partner intent and capability. Various modalities exist to convey such information between human agents, however, natural interactions between humans and robots are difficult to characterise and achieve. To enhance inter-agent communication, predictive models for human movement have been devised. One such model is Fitts' law. Many works using Fitts' law rely on massless interfaces. However, this coupling between human and robot, and the inertial effects experienced, may affect the predictive ability of Fitts' law. Experiments were conducted on human-robot dyads during a target-directed force exertion task. From the interactions, the results indicate that there is no observable effect regarding Fitts' law's predictive ability.
Thiyagarajan, K, Acharya, P, Piyathilaka, L & Kodagoda, S 1970, 'Numerical Modeling of the Effects of Electrode Spacing and Multilayered Concrete Resistivity on the Apparent Resistivity Measured Using Wenner Method', 2020 15th IEEE Conference on Industrial Electronics and Applications (ICIEA), 2020 15th IEEE Conference on Industrial Electronics and Applications (ICIEA), IEEE, Kristiansand, Norway, pp. 200-206.
View/Download from: Publisher's site
Thiyagarajan, K, Kodagoda, S & Ulapane, N 1970, 'Short-term Time Series Forecasting of Concrete Sewer Pipe Surface Temperature', 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), IEEE, Shenzhen, China, pp. 1194-1199.
View/Download from: Publisher's site
Thiyagarajan, K, Kodagoda, S, Ulapane, N & Prasad, M 1970, 'A Temporal Forecasting Driven Approach Using Facebook’s Prophet Method for Anomaly Detection in Sewer Air Temperature Sensor System', 2020 15th IEEE Conference on Industrial Electronics and Applications (ICIEA), 2020 15th IEEE Conference on Industrial Electronics and Applications (ICIEA), IEEE, Kristiansand, Norway, pp. 25-30.
View/Download from: Publisher's site
To, KYC, Lee, JJH, Yoo, C, Anstee, S & Fitch, R 1970, 'Streamline-Based Control of Underwater Gliders in 3D Environments', 2019 IEEE 58th Conference on Decision and Control (CDC), 2019 IEEE 58th Conference on Decision and Control (CDC), IEEE, Nice, France, pp. 8303-8310.
View/Download from: Publisher's site
View description>>
Autonomous underwater gliders use buoyancy control to achieve forwardpropulsion via a sawtooth-like, rise-and-fall trajectory. Because gliders areslow-moving relative to ocean currents, glider control must consider the effectof oceanic flows. In previous work, we proposed a method to control underwatervehicles in the (horizontal) plane by describing such oceanic flows in terms ofstreamlines, which are the level sets of stream functions. However, the generalanalytical form of streamlines in 3D is unknown. In this paper, we show howstreamline control can be used in 3D environments by assuming a 2.5D model ofocean currents. We provide an efficient algorithm that acts as a steeringfunction for a single rise or dive component of the glider's sawtoothtrajectory, integrate this algorithm within a sampling-based motion planningframework to support long-distance path planning, and provide several examplesin simulation in comparison with a baseline method. The key to our method'scomputational efficiency is an elegant dimensionality reduction to a 1D controlregion. Streamline-based control can be integrated within varioussampling-based frameworks and allows for online planning for gliders incomplicated oceanic flows.
To, KYC, Yoo, C, Anstee, S & Fitch, R 1970, 'Distance and Steering Heuristics for Streamline-Based Flow Field Planning', Proceedings - IEEE International Conference on Robotics and Automation, IEEE International Conference on Robotics and Automation, IEEE, Paris, France, pp. 1867-1873.
View/Download from: Publisher's site
View description>>
Motion planning for vehicles under the influence of flow fields can benefitfrom the idea of streamline-based planning, which exploits ideas from fluiddynamics to achieve computational efficiency. Important to such planners is anefficient means of computing the travel distance and direction between twopoints in free space, but this is difficult to achieve in strong incompressibleflows such as ocean currents. We propose two useful distance functions inanalytical form that combine Euclidean distance with values of the streamfunction associated with a flow field, and with an estimation of the strengthof the opposing flow between two points. Further, we propose steeringheuristics that are useful for steering towards a sampled point. We evaluatethese ideas by integrating them with RRT* and comparing the algorithm'sperformance with state-of-the-art methods in an artificial flow field and inactual ocean prediction data in the region of the dominant East AustralianCurrent between Sydney and Brisbane. Results demonstrate the method'scomputational efficiency and ability to find high-quality paths outperformingstate-of-the-art methods, and show promise for practical use with autonomousmarine robots.
Ulapane, N, Thiyagarajan, K & Kodagoda, S 1970, 'Binary Spectrum Feature for Improved Classifier Performance', 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), 2020 16th International Conference on Control, Automation, Robotics and Vision (ICARCV), IEEE, Shenzhen, China, pp. 1117-1122.
View/Download from: Publisher's site
Ulapane, N, Thiyagarajan, K & Kodagoda, S 1970, 'Hyper-Parameter Initialization for Squared Exponential Kernel-based Gaussian Process Regression', 2020 15th IEEE Conference on Industrial Electronics and Applications (ICIEA), 2020 15th IEEE Conference on Industrial Electronics and Applications (ICIEA), IEEE, Kristiansand, Norway, pp. 1154-1159.
View/Download from: Publisher's site
Ulapane, N, Thiyagarajan, K & Kodagoda, S 1970, 'System Identification of Static Nonlinear Elements: A Unified Approach of Active Learning, Over-fit Avoidance, and Model Structure Determination', 2020 15th IEEE Conference on Industrial Electronics and Applications (ICIEA), 2020 15th IEEE Conference on Industrial Electronics and Applications (ICIEA), IEEE, Kristiansand, Norway, pp. 1001-1006.
View/Download from: Publisher's site
Usayiwevu, M, Le Gentil, C, Mehami, J, Yoo, C, Fitch, R & Vidal-Calleja, T 1970, 'Information Driven Self-Calibration for Lidar-Inertial Systems', 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, Las Vegas, NV, USA, pp. 9961-9967.
View/Download from: Publisher's site
View description>>
Multi-modal estimation systems have the advantage of increased accuracy and robustness. To achieve accurate sensor fusion with these types of systems, a reliable extrinsic calibration between each sensor pair is critical. This paper presents a novel self-calibration framework for lidar-inertial systems. The key idea of this work is to use an informative path planner to find the admissible path that produces the most accurate calibration of such systems in an unknown environment within a given time budget. This is embedded into a simultaneous localization, mapping and calibration lidar-inertial system, which involves challenges in dealing with agile motions for excitation and large amount of data. Our approach has two stages: firstly, the environment is explored and mapped following a pre-defined path; secondly, the map is exploited to find a continuous and differentiable path that maximises the information gain within a sampling-based planner. We evaluate the proposed self-calibration method in a simulated environment and benchmark it with standard predefined paths to show its performance.
Wang, T, Lu, W, Yan, Z & Liu, D 1970, 'DOB-Net: Actively Rejecting Unknown Excessive Time-Varying Disturbances', 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020 IEEE International Conference on Robotics and Automation (ICRA), IEEE, Paris, France, pp. 1881-1887.
View/Download from: Publisher's site
View description>>
This paper presents an observer-integrated Reinforcement Learning (RL) approach, called Disturbance OB-server Network (DOB-Net), for robots operating in environments where disturbances are unknown and time-varying, and may frequently exceed robot control capabilities. The DOB-Net integrates a disturbance dynamics observer network and a controller network. Originated from conventional DOB mechanisms, the observer is built and enhanced via Recurrent Neural Networks (RNNs), encoding estimation of past values and prediction of future values of unknown disturbances in RNN hidden state. Such encoding allows the controller generate optimal control signals to actively reject disturbances, under the constraints of robot control capabilities. The observer and the controller are jointly learned within policy optimization by advantage actor critic. Numerical simulations on position regulation tasks have demonstrated that the proposed DOB-Net significantly outperforms conventional feedback controllers and classical RL policy.
Yang, T, Valls Miro, J, Wang, Y & Xiong, R 1970, 'Non-revisiting Coverage Task with Minimal Discontinuities for Non-redundant Manipulators', Robotics: Science and Systems XVI, Robotics: Science and Systems 2020, Robotics: Science and Systems Foundation, ELECTR NETWORK.
View/Download from: Publisher's site
Yoo, C, Lensgraf, S, Fitch, R, Clemon, LM & Mettu, R 1970, 'Toward Optimal FDM Toolpath Planning with Monte Carlo Tree Search', Proceedings - IEEE International Conference on Robotics and Automation, pp. 4037-4043.
View/Download from: Publisher's site
View description>>
The most widely used methods for toolpath planning in fused deposition 3Dprinting slice the input model into successive 2D layers in order to constructthe toolpath. Unfortunately slicing-based methods can incur a substantialamount of wasted motion (i.e., the extruder is moving while not printing),particularly when features of the model are spatially separated. In recentyears we have introduced a new paradigm that characterizes the space offeasible toolpaths using a dependency graph on the input model, along withseveral algorithms to search this space for toolpaths that optimize objectivefunctions such as wasted motion or print time. A natural question that arisesis, under what circumstances can we efficiently compute an optimal toolpath? Inthis paper, we give an algorithm for computing fused deposition modeling (FDM)toolpaths that utilizes Monte Carlo Tree Search (MCTS), a powerfulgeneral-purpose method for navigating large search spaces that is guaranteed toconverge to the optimal solution. Under reasonable assumptions on printergeometry that allow us to compress the dependency graph, our MCTS-basedalgorithm converges to find the optimal toolpath. We validate our algorithm ona dataset of 75 models and show it performs on par with our previous best localsearch-based algorithm in terms of toolpath quality. In prior work wespeculated that the performance of local search was near optimal, and weexamine in detail the properties of the models and MCTS executions that lead tobetter or worse results than local search.
You, A, Sukkar, F, Fitch, R, Karkee, M & Davidson, JR 1970, 'An Efficient Planning and Control Framework for Pruning Fruit Trees', 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020 IEEE International Conference on Robotics and Automation (ICRA), IEEE, Paris, Texas (Virtual), pp. 3930-3936.
View/Download from: Publisher's site
View description>>
Dormant pruning is a major cost component of fresh market tree fruit production, nearly equal in scale to harvesting the fruit. However, relatively little focus has been given to the problem of pruning trees autonomously. In this paper, we introduce a robotic system consisting of an industrial manipulator, an eye-in-hand RGB-D camera configuration, and a custom pneumatic cutter. The system is capable of planning and executing a sequence of cuts while making minimal assumptions about the environment. We leverage a novel planning framework designed for high-throughput operation which builds upon previous work to reduce motion planning time and sequence cut points intelligently. In end-to-end experiments with a set of ten different branch configurations, the system achieved a high success rate in plan execution and a 1.5x speedup in throughput versus a baseline planner, representing a significant step towards the goal of practical implementation of robotic pruning.
Zhang, Y, Falque, R, Zhao, L, Huang, S & Hu, B 1970, 'Deep Learning Assisted Automatic Intra-operative 3D Aortic Deformation Reconstruction', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Springer International Publishing, pp. 660-669.
View/Download from: Publisher's site
View description>>
Endovascular interventions rely on 2D X-ray fluoroscopy for 3D catheter manipulation. The dynamic nature of aorta prevents the pre-operative CT/MRI data to be used directly as the live 3D guidance since the vessel deforms during the surgery. This paper provides a framework that reconstructs the live 3D aortic shape by fusing a 3D static pre-operative model and the 2D intra-operative fluoroscopic images. The proposed framework recovers aortic 3D shape automatically and computationally efficient. A deep learning approach is adopted as the front-end for extracting features from fluoroscopic images. A signed distance field based correspondence method is employed for avoiding the repeated feature-vertex matching while maintaining the correspondence accuracy. The warp field of 3D deformation is estimated by solving a non-linear least squares problem based on the embedded deformation graph. Detailed phantom experiments are conducted, and the results demonstrate the accuracy of the proposed framework as well as the potential clinical value of the technique.
Zhang, Y, Zhao, L & Huang, S 1970, 'Aortic 3D Deformation Reconstruction using 2D X-ray Fluoroscopy and 3D Pre-operative Data for Endovascular Interventions', 2020 IEEE International Conference on Robotics and Automation (ICRA), 2020 IEEE International Conference on Robotics and Automation (ICRA), IEEE, Paris, France (Virtual), pp. 2393-2399.
View/Download from: Publisher's site
View description>>
Current clinical endovascular interventions rely on 2D guidance for catheter manipulation. Although an aortic 3D surface is available from the pre-operative CT/MRI imaging, it cannot be used directly as a 3D intra-operative guidance since the vessel will deform during the procedure. This paper aims to reconstruct the live 3D aortic deformation by fusing the static 3D model from the pre-operative data and the 2D live imaging from fluoroscopy. In contrast to some existing deformation reconstruction frameworks which require 3D observations such as RGB-D or stereo images, fluoroscopy only presents 2D information. In the proposed framework, a 2D-3D registration is performed and the reconstruction process is formulated as a non-linear optimization problem based on the deformation graph approach. Detailed simulations and phantom experiments are conducted and the result demonstrates the reconstruction accuracy and robustness, as well as the potential clinical value of this framework.
