Showing papers presented at "Field and Service Robotics in 2008"

Autonomous fruit picking machine: A robotic apple harvester

Abstract: This paper describes the construction and functionality of an Autonomous Fruit Picking Machine (AFPM) for robotic apple harvesting. The key element for the success of the AFPM is the integrated approach which combines state of the art industrial components with the newly designed flexible gripper. The gripper consist of a silicone funnel with a camera mounted inside. The proposed concepts guarantee adequate control of the autonomous fruit harvesting operation globally and of the fruit picking cycle particularly. Extensive experiments in the field validate the functionality of the AFPM.

State space sampling of feasible motions for high-performance mobile robot navigation in complex environments

Abstract: Sampling in the space of controls or actions is a well-established method for ensuring feasible local motion plans. However, as mobile robots advance in performance and competence in complex environments, this classical motion-planning technique ceases to be effective. When environmental constraints severely limit the space of acceptable motions or when global motion planning expresses strong preferences, a state space sampling strategy is more effective. Although this has been evident for some time, the practical question is how to achieve it while also satisfying the severe constraints of vehicle dynamic feasibility. The paper presents an effective algorithm for state space sampling utilizing a model-based trajectory generation approach. This method enables high-speed navigation in highly constrained and-or partially known environments such as trails, roadways, and dense off-road obstacle fields. © 2008 Wiley Periodicals, Inc.

Modelling Smooth Paths Using Gaussian Processes

Abstract: A generative model based on the gaussian mixture model and gaussian processes is presented in this paper. Typical motion paths are learnt and then used for motion prediction using this model. The principal novel aspect of this approach is the modelling of paths using gaussian processes. It allows the representation of smooth trajectories and avoids discretization problems found in most existing methods. Gaussian processes not only provides a comprehensive and formal theoretical framework to work with, it also lends itself naturally to path clustering using gaussian mixture models. Learning is performed using expectation maximization where the E-Step uses variational methods to maximize its lower bound before optimization over parameters are performed in the M-Step.

Combined Visual and Inertial Navigation for an Unmanned Aerial Vehicle

Abstract: We describe an UAV navigation system which combines stereo visual odometry with inertial measurements from an IMU. Our approach fuses the motion estimates from both sensors in an extended Kalman filter to determine vehicle position and attitude. We present results using data from a robotic helicopter, in which the visual and inertial system produced a final position estimate within 1% of the measured GPS position, over a flight distance of more than 400 meters. Our results show that the combination of visual and inertial sensing reduced overall positioning error by nearly an order of magnitude compared to visual odometry alone.

Autonomous underground tramming for center-articulated vehicles

Abstract: This paper describes the design, implementation, and field testing of an infrastructureless system for autonomous tramming (or hauling) of a center-articulated underground mining vehicle. Such vehicles are ubiquitous in underground mining, and effective automation of their tramming function has been a sought-after technology for more than a decade. This paper reports on the successful development of a fast, reliable, and robust “autotramming” technology that does not require the installation of fixed infrastructure. Included are descriptions of the chosen control architecture, map-based localization technique, and the results of integration and field testing. © 2008 Wiley Periodicals, Inc.

Monte Carlo localization in outdoor terrains using multilevel surface maps

Abstract: We propose a novel combination of techniques for robustly estimating the position of a mobile robot in outdoor environments using range data. Our approach applies a particle filter to estimate the full six-dimensional state of the robot and utilizes multilevel surface maps, which, in contrast to standard elevation maps, allow the robot to represent vertical structures and multiple levels in the environment. We describe probabilistic motion and sensor models to calculate the proposal distribution and to evaluate the likelihood of observations. We furthermore describe an active localization approach that actively selects the sensor orientation of the two-dimensional laser range scanner to improve the localization results. To efficiently calculate the appropriate orientation, we apply a clustering operation on the particles and evaluate potential orientations on the basis of these clusters. Experimental results obtained with a mobile robot in large-scale outdoor environments indicate that our approach yields robust and accurate position estimates. The experiments also demonstrate that multilevel surface maps lead to a significantly better localization performance than standard elevation maps. They additionally show that further accuracy is obtained from the active sensing approach. © 2008 Wiley Periodicals, Inc.

Experiments in Autonomous Reversing of a Tractor-Trailer System

Abstract: Trailer reversing is a problem frequently considered in the literature, usually with fairly complex non-linear control theory based approaches. In this paper, we overview our simple method for stabilizing a tractor-trailer system to a trajectory based on the notion of controlling the hitch-angle of the trailer rather than the steering angle of the tractor. The performance of this control method, as implemented on the CSIRO Autonomous Tractor, is then experimentally compared against a number of human drivers, showing that this method performs as well as a moderately skilled human driver, even though the system is significantly handicapped in terms of steering actuation speed and by errors in localization.

Magnetic Wall Climbing Robot for Thin Surfaces with Specific Obstacles

Abstract: This paper describes a novel solution to a mobile climbing robot on magnetic wheels, designed for inspecting the interior surfaces in gas tanks made out of thin metal sheets. These surfaces were inaccessible by previous climbing robots due to the following restrictions:

Design and Validation of a Mobile Robot for Power Line Inspection and Maintenance

Abstract: This paper presents the LineScout Technology, a mobile teleoperated robot for power line inspection and maintenance. Optimizing several geometric parameters achieved a compact design that was successfully tested over many line configurations and obstacle sequences. An overview of the technology is presented, including a description of the control strategy, followed by a section focusing on key aspects of the prototype thorough validation. Working on live lines, up to 735 kV and 1,000 A, means that the technology must be robust to electromagnetic interference. The third generation prototype, tested in laboratory and in field conditions, is now ready to undertake inspection pilot projects.

Reactive Collision Avoidance for Unmanned Aerial Vehicles Using Doppler Radar

Abstract: Research into reactive collision avoidance for unmanned aerial vehicles has been conducted on unmanned terrestrial and mini aerial vehicles utilising active Doppler radar obstacle detection sensors. Flight tests conducted by flying a mini UAV at an obstacle have confirmed that a simple reactive collision avoidance algorithm enables aerial vehicles to autonomously avoid obstacles. This builds upon simulation work and results obtained using a terrestrial vehicle that had already confirmed that active sensors and a reactive collision avoidance algorithm are able to successfully find a collision free path through an obstacle field.

Update Policy of Dense Maps: Efficient Algorithms and Sparse Representation

Abstract: Providing a robot with a fully detailed map is one appealing key for the Simultaneous Localisation and Mapping (SLAM) problem. It gives the robot a lot of hints to solve either the data association or the localisation problem itself. The more details are in the map, the more chances are that different places may appear differently, solving ambiguities. The more landmarks are used, the more accurate are the algorithms that solve the localisation problem since in a least square sense an approximation of the solution is more precise. Last, it helps a lot in the presence of a few dynamic objects because these moving parts of the environment remain marginal in the amount of data used to model the map and can thus be filtered out. For instance, the moving objects can be detected or cancelled in the localisation procedure by robust techniques using Monte-Carlo algorithms [6] or RANSAC [4].

Tree Measurement and Simultaneous Localization and Mapping System for Forest Harvesters

Abstract: For the last decades, measurement and automation systems in Nordic cut-to-length forestry machines have evolved gradually. These heavy duty machines are lighter, faster and more accurate than ever before but the basic technologies and operation have remained the same. In many respects, their current automation systems have reached their limits. The Forestrix project studies how advances in mobile robotics could be applied in the field of forestry machine automation. Machine vision systems and scanning laser range finders have established themselves as standard equipment in mobile robotics. With the new sensor and computing technologies it is possible to get information about the surrounding forest, such as tree diameters, positions and stand density. This information can be used on-line in operator’s decision support system, or off-line in a forest asset management system. This paper describes the prototype measurement platform and the software algorithms developed in the Forestrix project. Results from tests with an all terrain vehicle are also presented.

Robust trajectory tracking for a reversing tractor trailer

Abstract: Tractor-trailer reversing is a classical nonlinear control problem in which many of the solutions proposed in the literature perform poorly in the presence of real-world constraints such as steering angle, rate limits, and lags. In this paper we describe a new method in which an inner loop controls the hitch angle of the trailer, creating a virtual articulated vehicle to which existing control techniques can be applied. We provide an analysis of the stability and convergence properties of this control approach, as well as experimental results that illustrate the robustness of this approach to model estimation errors, low-level control loop dynamics, and other disturbances introduced by, for example, state estimation errors. © 2008 Wiley Periodicals, Inc.

Teleoperation of a Surgical Robot Via Airborne Wireless Radio and Transatlantic Internet Links

Abstract: Robotic assisted surgery generates the possibility of remote operation between surgeon and patient. We need better understanding of the engineering issues involved in operating a surgical robot in remote locations and through novel communication links between surgeon and surgery site. This paper describes two recent experiments in which we tested the RAVEN, a new prototype surgical robot manipulation system, in field and laboratory conditions. In the first experiment, the RAVEN was deployed in a pasture and ran on generator power. Telecommunication with the surgical control station was provided by a novel airborne radio link supported by an unmanned aerial vehicle. In the second experiment, the RAVEN was teleoperated via Internet between Imperial College in London and the BioRobotics Lab at the University of Washington in Seattle. Data are reported on surgeon completion times for basic tasks and on network latency experience. The results are a small step towards teleoperated surgical robots which can be rapidly deployed in emergency situations in the field.

Robust vision-based underwater homing using self-similar landmarks

Abstract: Next-generation autonomous underwater vehicles (AUVs) will be required to robustly identify underwater targets for tasks such as inspection, localization, and docking. Given their often unstructured operating environments, vision offers enormous potential in underwater navigation over more traditional methods; however, reliable target segmentation often plagues these systems. This paper addresses robust vision-based target recognition by presenting a novel scale and rotationally invariant target design and recognition routine based on self-similar landmarks that enables robust target pose estimation with respect to a single camera. These algorithms are applied to an AUV with controllers developed for vision-based docking with the target. Experimental results show that the system performs exceptionally on limited processing power and demonstrates how the combined vision and controller system enables robust target identification and docking in a variety of operating conditions. © 2008 Wiley Periodicals, Inc.

Robust Feature Extraction and Matching for Omnidirectional Images

Abstract: This paper presents a new and robust method for extracting and matching visual vertical features between images taken by an omnidirectional camera. Matching robustness is achieved by creating a descriptor which is unique and distinctive for each feature. Furthermore, the proposed descriptor is invariant to rotation. The robustness of the approach is validated through real experiments with a wheeled robot equipped with an omnidirectional camera. We show that vertical lines are very well extracted and tracked during the robot motion. At the end, we also present an application of our algorithm to the robot simultaneous localization and mapping in an unknown environment.

Robust Vision-Based Underwater Target Identification and Homing Using Self-Similar Landmarks

Abstract: Next generation Autonomous Underwater Vehicles (AUVs) will be required to robustly identify underwater targets for tasks such as inspection, localisation and docking. Given their often unstructured operating environments, vision offers enormous potential in underwater navigation over more traditional methods, however, reliable target segmentation often plagues these systems. This paper addresses robust vision-based target recognition by presenting a novel scale and rotationally invariant target design and recognition routine based on Self-Similar Landmarks (SSL) that enables robust target pose estimation with respect to a single camera. These algorithms are applied to an AUV with controllers developed for vision-based docking with the target. Experimental results show that system performs exceptionally on limited processing power and demonstrates how the combined vision and controller systems enables robust target identification and docking in a variety of operating conditions.

A Feature Based Navigation System for an Autonomous Underwater Robot

Abstract: We present a system for autonomous underwater navigation as implemented on a Nekton Ranger autonomous underwater vehicle, AUV. This is one of the first implementations of a practical application for simultaneous localization and mapping on an AUV. Besides being an application of real-time SLAM, the implemtation demonstrates a novel data fusion solution where data from 7 sources are fused at different time scales in 5 separate estimators. By modularizing the data fusion problem in this way each estimator can be tuned separately to provide output useful to the end goal of localizing the AUV, on an a priori map. The Ranger AUV is equipped with a BlueView blazed array sonar which is used to detect features in the underwater environment. Underwater testing results are presented. The features in these tests are deployed radar reflectors.

Assessing Map Quality Using Conditional Random Fields

Abstract: This paper is concerned with assessing the quality of work-space maps. While there has been much work in recent years on building maps of field settings, little attention has been given to endowing a machine with introspective competencies which would allow assessing the reliability/plausibility of the representation. We classify regions in 3D point-cloud maps into two binary classes — “plausible” or “suspicious”. In this paper we concentrate on the classification of urban maps and use a Conditional Random Fields to model the intrinsic qualities of planar patches and crucially, their relationship to each other. A bipartite labelling of the map is acquired via application of the Graph Cut algorithm. We present results using data gathered by a mobile robot equipped with a 3D laser range sensor while operating in a typical urban setting.

State Space Sampling of Feasible Motions for High Performance Mobile Robot Navigation in Highly Constrained Environments

Abstract: Sampling in the space of controls or actions is a well-established method for ensuring feasible local motion plans. However, as mobile robots advance in performance and competence in complex outdoor environments, this classical motion planning technique ceases to be effective. When environmental constraints severely limit the space of acceptable motions or when global motion planning expresses strong preferences, a state space sampling strategy is more effective. While this has been clear for some time, the practical question is how to achieve it while also satisfying the severe constraints of vehicle dynamic feasibility. This paper presents an effective algorithm for state space sampling based on a model-based trajectory generation approach. This method enables high-speed navigation in highly constrained and/or partially known environments such as trails, roadways, and dense off-road obstacle fields.

Picking up an Unknown Object through Autonomous Modeling and Grasp Planning by a Mobile Manipulator

Abstract: This paper describes a novel framework for object picking and carrying task by a mobile manipulator. Conventionally, researches on mobile manipulator cope well with object manipulation task with utilizing predefined knowledge or specific tools. So these researches have an essential problem that a new target object cannot be added without relatively many preparation. On the other hand, in our framework of a robot system, because the robot can create the 3D shape model of the object and can plan a grasp pose for the object autonomously under the condition that only the position of the object is given to the robot. Experimental results show the effectiveness of our robot system to be implemented our proposed method.

Data-driven identification of group dynamics for motion prediction and control

Abstract: A distributed model structure for representing groups of coupled dynamic agents is proposed, and the least-squares method is used for fitting model parameters based on measured position data. The difference equation model embodies a minimalist approach, incorporating only factors essential to the movement and interaction of physical bodies. The model combines effects from an agent's inertia, interactions between agents, and interactions between each agent and its environment. Global positioning system tracking data were collected in field experiments from a group of 3 cows and a group of 10 cows over the course of several days using custom-designed, head-mounted sensor boxes. These data are used with the least-squares method to fit the model to the cow groups. The modeling technique is shown to capture overall characteristics of the group as well as attributes of individual group members. Applications to livestock management are described, and the potential for surveillance, prediction, and control of various kinds of groups of dynamic agents are suggested. © 2008 Wiley Periodicals, Inc.

Coordinated Search in Cluttered Environments Using Range from Multiple Robots

Abstract: In this paper, we describe real-time methods for incorporating non-line-of-sight range measurements into a framework for finding a non-adversarial target in cluttered environments using multiple robotic searchers. We extend previous coordinated search strategies to utilize information from noisy non-line-of-sight range measurements. Sensors using ultra-wideband radio are becoming available that provide range measurements to targets even when they are occluded. We present two Bayesian methods for updating the expected location of a mobile target and integrating these updates into planning. We present simulated results in a complex museum environment as well as on mobile robots. Our results show the success of our algorithms at utilizing information from measurements in a coordinated search framework.

The Small Outdoor Rover MERLIN and Its Assistance System for Tele-operations

Abstract: Tele-operated rovers offer huge application potential in the areas of emergency support, surveillance and security, in particular if they are small and easy to transport. In this context, the MERLIN outdoor rovers have been designed for robust operations in harsh outdoor environments at a total mass below 20 kg. The MERLIN rover system design aspects are described, including the rover on-board data processing and the sensor configuration. Special emphasis is on the remote operations assistance system, composed of the human-machine interface for appropriate presentation of relevant data at the tele-operator workplace and the autonomous reaction capabilities coordinated by the on-board data handling system.

Development of a Networked Robotic System for Disaster Mitigation

Abstract: In this paper, the research project named “networked robotic system for disaster mitigation” is introduced The project aims to develop key technologies for multiple robots to be teleoperated through a wireless communication network, which includes a satellite communication link, for surveillance tasks at a disaster site The robotic system consists of a large-scale outdoor robot and a group of smaller indoor robots The large-scale robot will serve as a carrier for the smaller robots which are deployed inside a partly-collapsed building A three-dimensional range sensor and an omnidirectional camera are used as tools to ease the teleoperation for the human operator This paper presents the mission scenario and the development status of the proposed networked multi-robotic system The results of the performance tests are also reported

A FastSLAM algorithm based on the unscented filtering with adaptive selective resampling

Abstract: A FastSLAM approach to the SLAM problem is considered in this paper. An improvement to the classical FastSLAM algorithm has been obtained by replacing the Extended Kalman Filters used in the prediction step and in the feature update with Unscented Kalman Filters and by introducing an adaptive selective resampling. The simulations confirm the effectiveness of the proposed modifications.

Experiments in Navigation and Mapping with a Hovering AUV

Abstract: This paper describes the basic control, navigation, and mapping methods and experiments a hovering autonomous underwater vehicle (AUV) designed to explore flooded cenotes in Mexico as part of the DEPTHX project. We describe the low level control system of the vehicle, and present a dead reckoning navigation filter that compensates for frequent Doppler velocity log (DVL) dropouts. Sonar data collected during autonomous excursions in a limestone quarry are used to generate a map of the quarry geometry.

Online Learning of Multiple Perceptual Models for Navigation in Unknown Terrain

Abstract: Autonomous robots in unknown and unstructured environments must be able to distinguish safe and unsafe terrain in order to navigate effectively. Stereo depth data is effective in the near field, but agents should also be able to observe and learn perceptual models for identifying traversable surfaces and obstacles in the far field. As the robot passes through the environment however, the appearance of ground plane and obstacles may vary, for example in open fields versus tree cover or paved versus gravel or dirt tracks. In this paper we describe a working robot navigation system based primarily on colour imaging, which learns sets of fast, efficient density-based models online. As the robot moves through the environment the system chooses whether to apply current models, discard inappropriate models or acquire new ones. These models operate on complex natural images and are acquired and used in real time as the robot navigates.

An Advanced Teleoperation Testbed

Abstract: Due to the technology available, most previous work in teleoperated robotics used relatively low-resolution video links and provided limited perceptual feedback to the teleoperator. In most cases, these projects reported only limited teleoperator success compared to vehicles with human drivers on-board. We set out to build a high-fidelity teleoperation system which takes advantage of recent technological advances. This system permits highly capable teleoperation and has allowed us to begin to investigate the minimum system requirements for effective teleoperation.

Robot Position Estimation on a RFID-Tagged Smart Floor

Abstract: We describe a navigation and position estimation system for mobile service robots which is based on RFID technology. This system avoids deficiencies of existing solutions and offers high flexibility and accuracy at moderate cost. The core idea of the system is to structure the work space by means of a Smart Floor in a manner which enables and supports reliable navigation and positioning with absolute accuracy over large distances. The “intelligence” of the floor is in a dense, area-wide network of thousands of RFID transponder, which are mounted underneath the regular floor covering and quasi serve as radio beacons. The system has been presented at CeBIT 2006 in Hannover by invitation of the German Ministry of Education and Research. InMach Intelligente Maschinen has further received the Walter Reis Innovation Award for Service Robotics for this innovative solution. It is important to note that although the primary focus of the work described in this paper is on the guidance of service robots in large-scale indoor environments, the Smart Floor described below provides the infrastructure for all kinds of location-based services which might be of relevance in a large facility. The economic value of these services may even exceed the economic value of automated cleaning.