Can supervise: YES
Hodges, J, Attia, T, Arukgoda, J, Kang, C, Cowden, M, Doan, L, Ranasinghe, R, Abdelatty, K, Dissanayake, G & Furukawa, T 2019, 'Multistage bayesian autonomy for high-precision operation in a large field', Journal of Field Robotics, vol. 36, no. 1, pp. 183-203.View/Download from: UTS OPUS or Publisher's site
© 2018 Wiley Periodicals, Inc. This paper presents a generalized multistage bayesian framework to enable an autonomous robot to complete high-precision operations on a static target in a large field. The proposed framework consists of two multistage approaches, capable of dealing with the complexity of high-precision operation in a large field to detect and localize the target. In the multistage localization, locations of the robot and the target are estimated sequentially when the target is far away from the robot, whereas these locations are estimated simultaneously when the target is close. A level of confidence (LOC) for each detection criterion of a sensor and the associated probability of detection (POD) of the sensor are defined to make the target detectable with different LOCs at varying distances. Differential entropies of the robot and target are used as a precision metric for evaluating the performance of the proposed approach. The proposed multistage observation and localization approaches were applied to scenarios using an unmanned ground vehicle (UGV) and an unmanned aerial vehicle (UAV). Results with the UGV in simulated environments and then real environments show the effectiveness of the proposed approaches to real-world problems. A successful demonstration using the UAV is also presented.
Thiyagarajan, K, Kodagoda, S, Ranasinghe, R, Vitanage, D & Iori, G 2018, 'Robust sensing suite for measuring temporal dynamics of surface temperature in sewers', Nature - Scientific Reports, vol. 8.View/Download from: Publisher's site
Thiyagarajan, K, Kodagoda, S, Nguyen, LV & Ranasinghe, R 2018, 'Sensor Failure Detection and Faulty Data Accommodation Approach for Instrumented Wastewater Infrastructures', IEEE Access, vol. 6, no. 1, pp. 56562-56562.View/Download from: UTS OPUS or Publisher's site
Nguyen, LV, Kodagoda, S, Ranasinghe, R & Dissanayake, G 2017, 'Adaptive Placement for Mobile Sensors in Spatial Prediction under Locational Errors', IEEE Sensors Journal, vol. 17, no. 3, pp. 794-802.View/Download from: UTS OPUS or Publisher's site
Dantanarayana, L, Dissanayake, G & Ranasinge, R 2016, 'C-LOG: A Chamfer distance based algorithm for localisation in occupancy grid-maps', CAAI Transactions on Intelligence Technology, vol. 1, no. 3, pp. 272-284.View/Download from: UTS OPUS or Publisher's site
Nguyen, LV, Kodagoda, S & Ranasinghe, R 2016, 'Spatial Sensor Selection via Gaussian Markov Random Fields', IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS, vol. 46, no. 9, pp. 1226-1239.View/Download from: UTS OPUS or Publisher's site
Nguyen, LV, Kodagoda, S, Ranasinghe, R & Dissanayake, G 2016, 'Information-Driven Adaptive Sampling Strategy for Mobile Robotic Wireless Sensor Network', IEEE Transactions on Control Systems Technology, vol. 24, no. 1, pp. 372-379.View/Download from: UTS OPUS or Publisher's site
This brief addresses the issue of monitoring physical spatial phenomena of interest using information collected by a resource-constrained network of mobile, wireless, and noisy sensors that can take discrete measurements as they navigate through the environment. We first propose an efficient novel optimality criterion for designing a sampling strategy to find the most informative locations in taking future observations to minimize the uncertainty at all unobserved locations of interest. This solution is proven to be within bounds. The computational complexity of this proposition is shown to be practically feasible. We then prove that under a certain condition of monotonicity property, the approximate entropy at resulting locations obtained by our proposed algorithm is within 1-(1/e) of the optimum, which is then utilized as a stopping criterion for the sampling algorithm. The criterion enables the prediction results to be within user-defined accuracies by controlling the number of mobile sensors. The effectiveness of the proposed method is illustrated using a prepublished data set.
Unicomb, J, Ranasinghe, RS, Dantanarayana, L & Dissanayake, G 2018, 'A Monocular Indoor Localiser based on an Extended Kalman Filter and Edge Images from a Convolutional Neural Network', International Conference on Intelligent Robots, IEEE, Madrid.View/Download from: UTS OPUS or Publisher's site
The main contribution of this paper is an extended Kalman filter (EKF)based algorithm for estimating the 6 DOF pose of a camera using monocular images of an indoor environment. In contrast to popular visual simultaneous localisation and mapping algorithms, the technique proposed relies on a pre-built map represented as an unsigned distance function of the ground plane edges. Images from the camera are processed using a Convolutional Neural Network (CNN)to extract a ground plane edge image. Pixels that belong to these edges are used in the observation equation of the EKF to estimate the camera location. Use of the CNN makes it possible to extract ground plane edges under significant changes to scene illumination. The EKF framework lends itself to use of a suitable motion model, fusing information from any other sensors such as wheel encoders or inertial measurement units, if available, and rejecting spurious observations. A series of experiments are presented to demonstrate the effectiveness of the proposed technique.
Tran, A, Liu, D, Ranasinghe, R & Carmichael, M 2018, 'Identifying Human Hand Orientation around a Cylindrical Handlebar for physical Human-Robot Interaction', International Symposium on Robotics, Munich, pp. 427-434.View/Download from: UTS OPUS
Tran, A, Liu, D, Ranasinghe, R & Carmichael, M 2018, 'Method for Quantifying a Robot's Confidence in its Human Co-worker in Human-Robot Cooperative Grit-Blasting', International Symposium on Robotics, Munich, pp. 474-481.View/Download from: UTS OPUS
Perera, K, Ranasinghe, R & Dissanayake, G 2017, 'A Neural Network Based Place Recognition Technique for a Crowded Indoor Environment', The 13th IEEE conference on industrial electronics and applications, IEEE, Siem Reap, Cambodia, pp. 1937-1942.View/Download from: UTS OPUS or Publisher's site
Place recognition in a crowded and cluttered environment is a challenging task due to its dynamic characteristics such as moving obstacles, varying lighting conditions and occlusions. This work presents a robust place recognition technique that could be applied into a similar environment, by combining well known Bag of Words technique with a feedforward neural network. The feedforward neural network we use have three layers with a single hidden layer and it relies on rectifier and softmax activation functions. We employ cross entropy function to model the cost of our neural network and utilize Adam algorithm for minimizing this cost at the training phase. The output layer with softmax activation in the neural network, produces a vector of probabilities which represent the likelihood of test image being captured from a given region. These values are further improved by incorporating a transition matrix which is based on the building layout. We have evaluated our neural network based place recognition technique with data collected from a crowded indoor shopping mall and promising results have been observed by this approach. We also have analyzed the behavior of neural network for changes in hyper-parameters and presented the results.
Wu, K, Li, X, Ranasinghe, R, Dissanayake, G & Liu, Y 2017, 'RISAS: A novel rotation, illumination, scale invariant appearance and shape feature', Proceedings - IEEE International Conference on Robotics and Automation, International Conference on Robotics and Automation, IEEE, Singapore, Singapore, pp. 4008-4015.View/Download from: UTS OPUS or Publisher's site
© 2017 IEEE. This paper presents a novel appearance and shape feature, RISAS, which is robust to viewpoint, illumination, scale and rotation variations. RISAS consists of a keypoint detector and a feature descriptor both of which utilise texture and geometric information present in the appearance and shape channels. A novel response function based on the surface normals is used in combination with the Harris corner detector for selecting keypoints in the scene. A strategy that uses the depth information for scale estimation and background elimination is proposed to select the neighbourhood around the keypoints in order to build precise invariant descriptors. Proposed descriptor relies on the ordering of both grayscale intensity and shape information in the neighbourhood. Comprehensive experiments which confirm the effectiveness of the proposed RGB-D feature when compared with CSHOT  and LOIND are presented. Furthermore, we highlight the utility of incorporating texture and shape information in the design of both the detector and the descriptor by demonstrating the enhanced performance of CSHOT and LOIND when combined with RISAS detector.
Unicomb, J, Dantanarayana, L, Arukgoda, J, Ranasinghe, R, Dissanayake, G & Furukawa, T 2017, 'Distance function based 6DOF localization for unmanned aerial vehicles in GPS denied environments', 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE, Vancouver, BC, Canada.View/Download from: UTS OPUS or Publisher's site
This paper presents an algorithm for localizing an unmanned aerial vehicle (UAV) in GPS denied environments. Localization is performed with respect to a pre-built map of the environment represented using the distance function of a binary mosaic, avoiding the need for extraction and explicit matching of visual features. Edges extracted from images acquired by an on-board camera are projected to the map to compute an error metric that indicates the misalignment between the predicted and true pose of the UAV. A constrained extended Kalman filter (EKF) framework is used to generate an estimate of the full 6-DOF location of the UAV by enforcing the condition that the distance function values are zero when there is no misalignment. Use of an EKF also makes it possible to seamlessly incorporate information from any other system on the UAV, for example, from its auto-pilot, a height sensor or an optical flow sensor. Experiments using a hexarotor UAV both in a simulation environment and in the field are presented to demonstrate the effectiveness of the proposed algorithm.
Perera, A, Arukgoda, J, Ranasinghe, RS & Dissanayake, G 2017, 'Localization System for Carers to Track Elderly People in Visits to a Crowded Shopping Mall', 2017 International Conference on Indoor Positioning and Indoor Navigation (IPIN), International Conference on Indoor Positioning and Indoor Navigation, IEEE, Sapporo, Japan, pp. 1-8.View/Download from: UTS OPUS or Publisher's site
This work presents a real-time localization system
developed for professional care givers to track residents of an
aged care facility during their visits to a crowded, multi-story
shopping mall. The proposed system consists of a Wi-Fi based
self-localization platform integrated into a wheeled walking frame
and an application installed in a hand-held tablet device for
displaying the locations of walker users. The density of people in
the shopping mall changes significantly during the day thus the
expected Wi-Fi signal strength at a given location is subject to
large variations. However, Identifying the location to be within
a given area is adequate and the average speed of motion is
less than 0.5 m/sec. In this paper, an algorithm that addresses
these unique requirements is presented. We exploit the signal
strength characteristics of existing Wi-Fi network and prior
knowledge of the building floor plans for developing our core
algorithm. The environments is divided in to cells that are
either enclosed spaces or divisions of larger open regions. The
probability density function of the Wi-Fi signal strength of each
cell is estimated using Kernel Density Estimation (KDE) and is
used in a probabilistic framework to estimate the user location.
Motion model of the users as well as the detection of floor
transition events are used to enhance the performance of the
location estimator. The algorithm was implemented using an
Odroid-C1 computer and a tablet with Android operating system.
Results obtained during field trials at Roselands Shopping Mall
in Sydney are presented.
Ranasinghe, R, Dissanayake, G, Furukawa, T, Arukgoda, J & Dantanarayana, L 2017, 'Environment Representation for Mobile Robot Localisation (Invited Paper)', 2017 IEEE International Conference on Industrial and Information Systems (ICIIS), IEEE International Conference on Industrial and Information Systems, IEEE, Peradeniya, Sri Lanka, pp. 296-301.View/Download from: UTS OPUS or Publisher's site
An adequate representation of the environment is an essential component of a mobile robot navigation system. This paper reviews the techniques reported in the literature for capturing the geometry of the space surrounding a mobile robot. In particular, the use of distance functions that combine some of the advantages of feature based and occupancy grid based representations for mobile robot localisation is described in detail. The effectiveness of various distance function based representations is demonstrated using a number of practical examples for localising ground and air vehicles.
Arukgoda, J, Ranasinghe, R, Dantanarayana, L, Dissanayake, G & Furukawa, T 2017, 'Vector Distance Function Based Map Representation for Robot Localisation', Australasian Conference on Robotics and Automation, Australasian Conference on Robotics and Automation, ACRA, Sydney Australia, pp. 1-8.View/Download from: UTS OPUS
This paper introduces the use of the vector
distance function (VDF) for representing environments,
particularly for the use in localisation
algorithms. It is shown that VDF has
a continuous derivative at the object boundary
in contrast to unsigned distance transform,
and does not require an environment populated
with closed object as in the case of the
signed distance transforms, the two most common
strategies reported in the literature for
representing environments based on distances
to nearest occupied regions. As such VDF overcomes
the main disadvantages of the existing
distance transform based representations in the
context of robot localisation. The key properties
of VDF are demonstrated and the use of
VDF in robot localisation using an optimization
based algorithm is illustrated using three
examples. It is shown that the proposed environment
representation and the localisation
algorithm is effective in providing accurate location
estimates as well as the associated uncertainties
Arukgoda, J, Ranasinghe, R, Danthanarayana, L, Dissanayake, G & Furukawa, T 2017, 'Vector distance function based map representation for robot localisation', Australasian Conference on Robotics and Automation, ACRA, pp. 165-172.
This paper introduces the use of the vector distance function (VDF) for representing environments, particularly for the use in localisation algorithms. It is shown that VDF has a continuous derivative at the object boundary in contrast to unsigned distance transform, and does not require an environment populated with closed object as in the case of the signed distance transforms, the two most common strategies reported in the literature for representing environments based on distances to nearest occupied regions. As such VDF overcomes the main disadvantages of the existing distance transform based representations in the context of robot localisation. The key properties of VDF are demonstrated and the use of VDF in robot localisation using an optimization based algorithm is illustrated using three examples. It is shown that the proposed environment representation and the localisation algorithm is effective in providing accurate location estimates as well as the associated uncertainties.
Ranasinghe, R & Kodagoda, S 2016, 'Spatial Prediction in Mobile Robotic Wireless Sensor Networks withNetwork Constraints', IEEE, International Conference on Control, Automation, Robotics and Vision, IEEE, Phuket, Thailand.View/Download from: UTS OPUS or Publisher's site
Furukawa, T, Takami, K, Tong, X, Watman, D, Hamed, A, Ranasinghe, R & Dissanayake, G 2015, 'Map-based navigation of an autonomous car using grid-based scan-to-map matching', Proceedings of the ASME Design Engineering Technical Conference, ASME International Design Engineering Technical Conferences & Computers and Information in Engineering Conference (IDETC/CIE), ASME, Boston, Massachusetts, USA, pp. 1-10.View/Download from: UTS OPUS or Publisher's site
© Copyright 2015 by ASME. This paper presents the map-based navigation of a car with autonomous capabilities using grid-based scan-to-map matching. The autonomous car used for demonstration is built based on Toyota Prius and can control the throttle, the brake and the steering by a computer. The proposed grid-based scan-to-map matching method represents a map with a finite number of grid cells, represents a scan and the map with scan points at each grid as normal distributions (NDs) and constructs a map by matching the scan NDs to the map NDs. The proposed method enables scan-based mapping at high speed while maintaining high accuracy. The representation of a grid cell of a map in terms of multiple NDs further enhances speed and accuracy. The accuracy analysis of the proposed method shows that a small robot with a wheel diameter of 8cm had yielded no loop closure error after the travel of 186m while the terminal position error by the GMapping was approximately 1m with the error growth of 1%. The application of the proposed method with the autonomous car has then demonstrated the ability of the proposed method for autonomous driving with varying and high speed and has also quantified the significance of speed for successful mapping in autonomous driving.
Wu, K, Dissanayake, G & Ranasinghe, R 2015, 'Active Recognition and Pose Estimation of Household Objects in Clutter', Proceedings of 2015 IEEE International Conference on Robotics and Automation (ICRA), IEEE International Conference on Robotics and Automation, IEEE, Seattle, pp. 4230-4237.View/Download from: UTS OPUS or Publisher's site
This paper presents an active object recognition and pose estimation system for household objects in a highly cluttered environment. A sparse feature model, augmented with the characteristics of features when observed from different viewpoints is used for recognition and pose estimation while a dense point cloud model is used for storing geometry. This strategy makes it possible to accurately predict the expected information available during the Next-Best-View planning process as both the visibility as well as the likelihood of feature matching can be considered simultaneously. Experimental evaluations of the object recognition and pose estimation with an RGB-D sensor mounted on a Turtlebot are presented.
Tran, A, Liu, D, Ranasinghe, R, Carmichael, M & Liu, C 2015, 'Analysis of Human grip strength in physical Human Robot Interaction', Proceedings - Analysis of Human grip strength in physical Human Robot Interaction, Conference on Applied Human Factors and Ergonomics, ELSEVIER SCIENCE BV, Las Vegas.View/Download from: UTS OPUS or Publisher's site
The purpose of this paper is to explore how an operator's grip plays a role in physical Human Robot Interaction (pHRI). By considering how the operator reacts to or initiates changes in control, it is possible to study the operator's grip pattern. By analyzing the grip pattern, it is possible to incorporate their natural response in order to create safer and more intuitive interfaces. An experiment where an exoskeleton and human collaborate in order to complete a path following task has been chosen to observe the forces applied by the user at the handle to determine the interaction between the operator and robot. A ThruMode Matrix Array sensor has been wrapped around the robot's handle to measure the applied pressure. By introducing the sensor it not only enables the measurement of the applied forces and how they are applied but also a measure of how tight the user is gripping the handle. Previous studies show that the natural response of a human to an unexpected event is to tighten their grip, indicating that how an operator grasps the handle can be related to the operator's intention. In order to investigate how the operator's grip of the handle changes, the experiments presented in this paper examine two different scenarios which might occur during an interaction, the first where the robot attempts to deviate from the path and the second where the operator wishes to deviate to a new path. The results of the experiments show that whether the operator or the robot initiates the transition, a measurable change in how the operator grasps the handle. The information in this paper can lead to new applications in pHRI by exploring the possible uses of an operator's grasping strength.
Furukawa, T, Dantanarayana, LI, Ziglar, J, Ranasinghe, R & Dissanayake, G 2015, 'Fast Global Scan Matching for High-Speed Vehicle Navigation', IEEE Xplore, IEEE International conference on Multisensor Fusion and Integration for Intelligent Systems, IEEE, San Diego, CA, USA, pp. 37-42.View/Download from: UTS OPUS or Publisher's site
Dantanarayana, L, Dissanayake, G, Ranasinghe, R & Furukawa, T 2015, 'An extended Kalman filter for localisation in occupancy grid maps', 2015 IEEE 10th International Conference on Industrial and Information Systems (ICIIS), IEEE International Conference on Industrial and Information Systems, IEEE, Peradeniya, Sri Lanka, pp. 419-424.View/Download from: UTS OPUS or Publisher's site
Ranasinghe, R, Dantanarayana, L, Tran, A, Lie, S, Behrens, M & Liu, L 2014, 'Smart Hoist: An Assistive Robot to Aid Carers', Proceedings for the International Conference on Control Automation Robotics & Vision, International Conference on Control, Automation, Robotics and Vision, IEEE, Singapore, pp. 1285-1291.View/Download from: UTS OPUS or Publisher's site
Assistive Robotics(AR) is a rapidly expanding field, implementing advanced intelligent machines which are able to work collaboratively with a range of human users; as assistants, tools and as companions. These AR devices can assist stretched carers at residential aged care facilities to safely enhance their capacity and to improve the quality of care services. The research work presented in this paper describes the pre- liminary outcomes of a design, development and implementation of a patient lifting AR device (Smart Hoist) to reduce lower back injuries to carers while transferring patients in aged care facilities. The proposed solution, a modified conventional lifter device which consists of several sensors capable of interacting with the Smart Hoist operator and its environment, and a set of powered wheels. This solution helps carers to manoeuvre the Smart Hoist safely and intuitively. Preliminary results collected from an evaluation of the Smart Hoist conducted at the premises of IRT Woonona residential care facility confirm the improved safety, comfort and confidence for the carers.
Dantanarayana, L, Ranasinghe, R, Tran, A, Liu, D & Dissanayake, G 2014, 'A novel collaboratively designed robot to assist carers', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), International Conference on Social Robotics (ICSR), SPRINGER-VERLAG BERLIN, Sydney, Australia, pp. 105-114.View/Download from: UTS OPUS or Publisher's site
© Springer International Publishing Switzerland 2014. This paper presents a co-design process and an assisted navigation strategy that enables a novel assistive robot, Smart Hoist, to aid carers transferring non-ambulatory residents. Smart Hoist was codesigned with residents and carers at IRT Woonona residential care facility to ensure that the device can coexist in the facility, while providing assistance to carers with the primary aim of reducing lower back injuries, and improving the safety of carers and patients during transfers. The Smart Hoist is equipped with simple interfaces to capture user intention in order to provide assisted manoeuvring. Using the RGB-D sensor attached to the device, we propose a method of generating a repulsive force that can be combined with the motion controller's output to allow for intuitive manoeuvring of the Smart Hoist, while negotiating with the environment. Extensive user trials were conducted on the premises of IRTWoonona residential care facility and feedback from end users confirm its intended purpose of intuitive behaviour, improved performance and ease of use.
Kanzhi, WU, RANASINGHE, R & DISSANAYAKE, G 2014, 'A Fast Pipeline for Textured Object Recognition in Clutter using an RGB-D Sensor', Control Automation Robotics Vision (ICARCV), 2014 13th International Conference on, International Conference on Control, Automation, Robotics and Vision, IEEE, Marina Bay Sands, Singapore, pp. 1650-1655.View/Download from: UTS OPUS or Publisher's site
This paper presents a modular algorithm pipeline for recognizing textured household objects in cluttered environment and estimating 6 DOF poses using an RGB-D sensor. The method draws from recent advances in this area and introduces a number of innovations that enable improved performances and faster operational speed in comparison with the state-of-the-art. The pipeline consists of (i) support plane subtraction (ii) SIFT feature extraction and approximate nearest neighbour based
matching (iii) feature clustering using 3D Eculidean distances (iv) SVD based pose estimation in combination with a outlier rejection strategy named SORSAC ( Spatially ORdered RAndom Consensus ) and (v) a pose combination and refinement step to
combine overlapping identical instances and to refine the pose estimation result by removing incorrect hypothesis. Quantitative comparisons with the MOPED  system on self-constructed dataset are presented to demonstrate the effectiveness of the pipeline.
Nguyen, LV, Kodagoda, S, Ranasinghe, R & Dissanayake, G 2014, 'Mobile Robotic Wireless Sensor Networks for Efficient Spatial Prediction', 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), IEEE/RSJ International Conference on Intelligent Robots and Systems, IEEE, Chicago, IL, USA, pp. 1176-1181.View/Download from: UTS OPUS or Publisher's site
This paper addresses the issue of monitoring physical spatial phenomena of interest utilizing the information collected by a network of mobile, wireless and noisy sensors that can take discrete measurements as they navigate through the environment. The spatial phenomenon is statistically modelled by a Gaussian Markov Random Field (GMRF) with hyperparameters that are learnt as the measurements accumulate over
time. In this context, the GMRF approximately represents the spatial field on an irregular lattice of triangulation by exploiting a stochastic partial differential equation (SPDE) approach, which benefits remarkably in computation due to the sparsity
of the precision matrix. A technique of the one-step-ahead forecast is employed to predict the future measurements that are required to find the optimal sampling locations. It is shown that optimizing the sampling path problem with the logarithm
of the determinant either of a covariance matrix using a GP model or of a precision matrix using a GMRF model for mobile robotic wireless sensor networks (MRWSNs) even by a greedy algorithm is impractical. This paper proposes an efficient novel
optimality criterion for the adaptive sampling strategy to find the most informative locations in taking future observations that minimize the uncertainty at unobserved locations. The computational complexity of our proposed method is linear, which makes the MRWSN scalable and practically feasible. The effectiveness of the proposed approach is compared and demonstrated using a pre-published data set with appealing results.
Nguyen, LV, Kodagoda, S, Ranasinghe, R & Dissanayake, G 2014, 'Spatially-Distributed Prediction with Mobile Robotic Wireless Sensor Networks', 2014 13th International Conference on Control, Automation, Robotics & Vision, International Conference on Control, Automation, Robotics and Vision, Institute of Electrical and Electronics Engineers Inc., Marina Bay Sands, Singapore, pp. 1153-1158.View/Download from: UTS OPUS or Publisher's site
This paper presents a distributed spatial estimation and prediction approach to address the centrally-computed scheme of Gaussian Process regression at each robotic sensor in resource-constrained networks of mobile, wireless and noisy agents monitoring physical phenomena of interest. A mobile sensor independently estimate its own parameters using collective measurements from itself and local neighboring agents as they navigate through the environment. A spatially-distributed prediction algorithm is designed utilizing methods of Jacobi overrelaxation and discrete-time average consensus to enable a robotic sensor to update its estimation of obtaining the global model parameters and recursively compute the global goal of inference. A distributed navigation strategy is also considered to drive sensors to the most uncertain locations enhancing the quality of prediction and learning parameters. Experimental results in a real-world data set illustrate the effectiveness of the proposed approach and is highly comparable to those of the centralized scheme.
Nguyen, LV, Kodagoda, S, Ranasinghe, R, Dissanayake, G, Bustamante, H, Vitanage, D & Nguyen, T 2014, 'Spatial Prediction of Hydrogen Sulfide in Sewers with a Modified Gaussian Process Combined Mutual Information', 2014 13th International Conference on Control, Automation, Robotics & Vision, International Conference on Control, Automation, Robotics and Vision, Institute of Electrical and Electronics Engineers Inc., Marina Bay Sands, Singapore, pp. 1130-1135.View/Download from: UTS OPUS or Publisher's site
This paper proposes a data driven machine learning model for spatial prediction of hydrogen sulfide (H2S) in a gravity sewer system. The gaseous H2S in the overhead of the gravity sewer is modelled using a Gaussian Process with a new covariance function due to constraints of sewer boundaries. The covariance function is proposed based on the distance between two locations computed along the lengths of the sewer network. A mutual information based strategy is used to choose the best k sensor measurements and their locations from among n potential sensor observations and their locations. This provably NP-hard combinatorial sensor selection problem is addressed by maximizing the mutual information between the selected locations and the locations that are not selected or do not have any sensor deployments. A proof-of-concept study was carried out comparing the spatial prediction of H2S with a complex model currently used by Sydney Water. The proposed approach is shown to be effective in both modelling and predicting the H2S spatial concentrations in sewers as well as identifying optimal number of H2S sensors and their locations for a required level of prediction accuracy.
Dantanarayana, LI, Ranasinghe, R & Dissanayake, G 2013, 'C-LOG: A Chamfer Distance Based Method for Localisation in Occupancy Grid-maps', IEEE/RSJ International Conference on Intelligent Robots and Systems 2013, IEEE/RSJ International Conference on Intelligent Robots and Systems, IEEE, Tokyo, Japan, pp. 376-381.View/Download from: UTS OPUS or Publisher's site
In this paper, the problem of localising a robot within a known two-dimensional environment is formulated as one of minimising the Chamfer Distance between the corresponding occupancy grid map and information gathered from a sensor such as a laser range finder. It is shown that this nonlinear optimisation problem can be solved efficiently and that the resulting localisation algorithm has a number of attractive characteristics when compared with the conventional particle filter based solution for robot localisation in occupancy grids. The proposed algorithm is able to perform well even when robot odometry is unavailable, insensitive to noise models and does not critically depend on any tuning parameters. Experimental results based on a number of public domain datasets as well as data collected by the authors are used to demonstrate the effectiveness of the proposed algorithm.
Nguyen, LV, Kodagoda, S, Ranasinghe, R & Dissanayake, G 2013, 'Locational Optimization based Sensor Placement for Monitoring Gaussian Processes Modeled Spatial Phenomena', Proc. 2013 IEEE 8th Conference on Industrial Electronics and Applications, IEEE Conference on Industrial Electronics and Applications, IEEE, Melbourne, Australia, pp. 1-6.View/Download from: UTS OPUS or Publisher's site
This paper addresses the sensor placement problem associated with monitoring spatial phenomena, where mobile sensors are located on the optimal sampling paths yielding a lower prediction error. It is proposed that the spatial phenomenon to be monitored is modeled using a Gaussian Process and a variance based density function is employed to develop an expected-value function. A locational optimization based effective algorithm is employed to solve the resulting minimization of the expectedvalue function. We designed a mutual information based strategy to select the most informative subset of measurements effectively with low computational time. Our experimental results on realworld datasets have verified the superiority of the proposed approach.
Wang, S, Kodagoda, S & Ranasinghe, R 2012, 'Road Terrain Type Classification based on Laser Measurement System Data', Proceedings of the Australasian Conference on Robotics and Automation, ACRA, Australasian Conference on Robotics and Automation, ARAA, Wellington, New Zealand, pp. 1-6.View/Download from: UTS OPUS
For road vehicles, knowledge of terrain types is useful in improving passenger safety and comfort. The conventional methods are susceptible to vehicle speed variations and in this paper we present a method of using Laser Measurement System (LMS) data for speed independent road type classification. Experiments were carried out with an instrumented road vehicle (CRUISE), by manually driving on a variety of road terrain types namely Asphalt, Concrete, Grass, and Gravel roads at different speeds. A looking down LMS is used for capturing the terrain data. The range data is capable of capturing the structural differences while the remission values are used to observe anomalies in surface reflectance properties. Both measurements are combined and used in a Support Vector Machines Classifier to achieve an average accuracy of 95% on different road types.
Nguyen, LV, Kodagoda, S, Ranasinghe, R & Dissanayake, G 2012, 'Simulated Annealing Based Approach for Near-Optimal Sensor Selection in Gaussian Processes', Proc. 2012 IEEE International Conference on Control, Automation and Information Sciences, International Conference on Control, Automation and Information Sciences, IEEE, Ho Chi Minh City, Vietnam, pp. 142-147.View/Download from: UTS OPUS or Publisher's site
This paper addresses the sensor selection problem associated with monitoring spatial phenomena, where a subset of k sensor measurements from among a set of n potential sensor measurements is to be chosen such that the root mean square prediction error is minimised. It is proposed that the spatial phenomena to be monitored is modelled using a Gaussian Process and a simulated annealing based approximately heuristic algorithm is used to solve the resulting minimisation problem. The algorithm is shown to be computationally efficient and is illustrated using both indoor and outdoor environment monitoring scenarios. It is shown that, although the proposed algorithm is not guaranteed to find the optimum, it always provides accurate solutions for broad range real-world and computer generated datasets.
Nguyen, LV, Ranasinghe, R, Kodagoda, S & Dissanayake, G 2012, 'Sensor Selection Based Routing for Monitoring Gaussian Processes Modeled Spatial Phenomena', Australasian Conference on Robotics and Automation 2012, Australasian Conference on Robotics and Automation, The ACRA 2012 Organising Committee, Wellington, New Zealand, pp. 1-7.View/Download from: UTS OPUS
This paper addresses the trade-off between the sensing quality and the energy consumption in the wireless sensor network associated with monitoring spatial phenomena. We use a non-parametric Gaussian Process to model the spatial phenomena to be monitored and simulated annealing based approximately heuristic algorithm for sensor selection. Our novel Sensor Selection based Routing (SSR) algorithm uses this model to identify the most informative nodes, which gives the root mean square prediction error less than a specified threshold, to construct the minimal energy expended routing tree rooted at the sink. Our experiments have verified that the proposed computationally efficient SSR algorithm has significant advantages over conventional techniques.
Dissanayake, G, Huang, S, Wang, Z & Ranasinghe, R 2011, 'A Review of Recent Developments in Simultaneous Localization and Mapping', Proceedings of the 6th International Conference on Industrial and Information Systems, International Conference on Industrial and Information Systems, IEEE, Sri Lanka, pp. 477-482.View/Download from: UTS OPUS or Publisher's site
Simultaneous Localization and Mapping (SLAM) problem has been an active area of research in robotics for more than a decade. Many fundamental and practical aspects of SLAM have been addressed and some impressive practical solutions have been demonstrated. The aim of this paper is to provide a review of the current state of the research on feature based SLAM, in particular to examine the current understanding of the fundamental properties of the SLAM problem and associated issues with the view to consolidate recent achievements.
Shi, L, Kodagoda, S & Ranasinghe, R 2011, 'Fast Indoor Scene Classification Using 3D Point Clouds', Australasian Conference on Robotics and Automation 2011, Australasian Conference on Robotics and Automation, The ACRA 2011 Organising Committee, Monash University, Melbourne Australia, pp. 1-7.View/Download from: UTS OPUS
A representation of space that includes both geometric and semantic information enables a robot to perform high-level tasks in complex environments. Identifying and categorizing environments based on onboard sensors are essential in these scenarios. The Kinectâ¢, a 3D low cost sensor is appealing in these scenarios as it can provide rich information. The downside is the presence of large amount of information, which could lead to higher computational complexity. In this paper, we propose a methodology to efficiently classify indoor environments into semantic categories using Kinectâ¢ data. With a fast feature extraction method along with an efficient feature selection algorithm (DEFS) and, support vector machines (SVM) classifier, we could realize a fast scene classification algorithm. Experimental results in an indoor scenario are presented including comparisons with its counterpart of commonly available 2D laser range finder data.
Ranasinghe, RS, Andrew, LLH, Hayes, DA & Everitt, D 2001, 'Scheduling disciplines for multimedia WLANs: Embedded round robin and wireless dual queue', IEEE International Conference on Communications, pp. 1243-1248.View/Download from: Publisher's site
Wireless local area networks have developed into a promising solution to support advanced data services in untethered environments. Selection of an efficient packet-scheduling scheme is important for managing the bandwidth while satisfying QoS requirements of active sessions having diverse traffic characteristics. The key difficulty is the distributed nature of the queues in the uplink, resulting in the scheduler having to trade off polling greedy stations against wasting resources by polling potentially idle stations. In order to address this, we propose a novel scheduling scheme, "Embedded Round Robin", which dynamically classifies stations as "busy" and "clear". We then extend the recently proposed Dual Queue scheduling discipline to the case of wireless networks.
Ranasinghe, RS, Andrew, LLH & Everitt, D 1999, 'Distributed contention-free traffic scheduling in IEEE 802.11 multimedia networks', IEEE Workshop on Local and Metropolitan Area Networks, pp. 18-28.View/Download from: Publisher's site
© 1999 IEEE. Wireless local area networks are a promising solution to support advanced data services in mobile environments. The IEEE 802.11 wireless LAN standard is emerging as a mature technology to support delay sensitive network services. In order to support these services the standard has proposed the use of a polling scheme; however, existing polling schemes require high communication overheads or suffer from unfairness. In this paper, we propose a distributed fair queueing algorithm distributed deficit round robin, which is compatible with the 802.11 medium access control rules. Software simulation of this scheme shows that it can manage a heterogeneous mix of delay sensitive traffic.
Ranasinghe, RS, Andrew, LLH & Everitt, D 1999, 'Impact of polling strategy on capacity of 802.11 based wireless multimedia LANs', IEEE International Conference on Networks, ICON, pp. 96-103.View/Download from: Publisher's site
Wireless local area networks are a viable technology to support multimedia traffic. One of the prominent wireless local area network standards being adopted as a mature technology is the IEEE 802.11 standard. In wireless multimedia networks, mobile stations will be capable of generating a heterogeneous traffic mix and therefore it is crucial to devise an efficient bandwidth allocation scheme to satisfy the quality of service requirements of each traffic class. In this paper we present a distributed fair queuing scheme which is compatible with the 802.11 standard and can manage bandwidth allocation for delay-sensitive traffic. The performance of the proposed scheme is evaluated by simulation, showing that a distributed version of deficit round robin outperforms the standard round robin service discipline from a capacity viewpoint. © 1999 IEEE.