This paper deals with the problem of planning multiple agent movements through a mission area modeled as a graph. The agents undergo classic communication and temporal constraints, and the quantitative objective is the minimization of the team’s traversal makespan. Additional specificities make the problem a particularly complex routing one: on some nodes are associated durative and coordinated actions to perform, which can involve either the co-presence of several agents or time dependencies. Also, some agents are deployable and able to move on denser graphs: namely, aerial robots can take off and land on the ground vehicle at any planned position, and can fly above ground obstacles. We model the problem as a CSP and solve it with a network flow model. Results show the efficacy of the model and resolution scheme, which provides solutions with one or two orders of magnitude smaller time than a numerical temporal hierarchical planning model, with only a few percent loss of optimality.

A recent advance in vector field path following is the introduction of the Parametric Guiding Vector Field method. It allows for singularity-free vector fields with strong convergence guarantees, usable even for self-intersecting paths. However, the method requires significant gain tuning for practical use. In particular, for unbounded paths, the gains will inevitably become ill-suited for efficient path following. We propose a method to overcome this issue by introducing a dynamic step adaptation strategy, which provides additional normalization properties to the field. This allows the following of unbounded curves and reduces the number of gains to tune. The proposed improvements are verified in simulations using the PaparazziUAV software.

Trans-media robots, capable of operating across diverse environments, add significant complexity for multi-robot task allocation and planning problems. This paper introduces a novel approach to plan missions for such multi-robot systems, that addresses the associated specific complexities and constraints. It streamlines the overall mission planning process by decomposing it into tractable sub-problems, and addresses the issues of coalition formation, path planning, and task scheduling. It provides mission plans in very little computation time and allows to tackle large missions intractable by global planners, with negligible loss in plan optimality.

Degraded visual environments have strong impacts on the quality of LiDAR data. Experiments in artificial fog conditions show that noise points caused by water particles present various distance distributions which depend on visibility. This article introduces a mathematical framework based on Bayesian inference and Markov Chain Monte-Carlo sampling to infer optical visibility from point clouds. The visibility estimation is cast as a classification problem based on the identification of the distance distributions. Contrary to deep learning methods, our approach is model-based and focuses on the design of a full probabilistic framework, more comprehensible, which is critical for autonomous driving. Ultimately, the impact of the optical visibility on the probability of detection of standard targets is assessed, which can yield improvements on autonomous vehicles performances in adverse weather conditions.

Covariance estimation for the Iterative Closest Point (ICP) point cloud registration algorithm is essential for state estimation and sensor fusion purposes. We argue that a major source of error for ICP is in the input data itself, from the sensor noise to the scene geometry. Benefiting from recent developments in deep learning for point clouds, we propose a data-driven approach to learn an error model for ICP. We estimate covariances modeling data-dependent heteroscedastic aleatoric uncertainty, and epistemic uncertainty using a variational Bayesian approach. The system evaluation is performed on LiDAR odometry on different datasets, highlighting good results in comparison to the state of the art.

Uncertainty estimation of visual feature is essential for vision-based systems, such as visual navigation. We show that errors inherent to visual tracking, in particular using KLT tracker, can be learned using a probabilistic loss function to estimate the covariance matrix on each tracked feature position. The proposed system is trained and evaluated on synthetic data, as well as on real data, highlighting good results in comparison to the state of the art. The benefits of the tracking uncertainty estimates are illustrated for visual motion estimation.

We propose a decentralized approach that simultaneously allocates and decomposes high level tasks among various robots. The approach exploits HTN structures and algorithms, that are used within an auction-based allocation scheme, and aims at dealing with complex tasks with causal or temporal relations. The paper formalizes the approach, and depicts how HTN planning processes are used to estimate bids and distribute tasks. Results on a statistical series of coverage problems are presented and their performance is assessed through a comparison with a state of the art algorithm.

Following recent developments, this paper investigates the possibility to predict uncertainty models for monocular graph SLAM using topological features of the problem. An architecture to learn relative (i.e. inter-keyframe) uncertainty models using the resistance distance in the covisibility graph is presented. The proposed architecture is applied to simulated UAV coverage path planning trajectories and an analysis of the approaches strengths and shortcomings is provided.

This work presents the deployment of UAVs for the exploration of clouds, from the system architecture and simulation tests to a real-flight campaign and trajectory analyzes. Thanks to their small size and low altitude, light UAVs have proven to be adapted for in-situ cloud data collection. The short life time of the clouds and limited endurance of the planes require to focus on the area of maximum interest to gather relevant data. Based on previous work on cloud adaptive sampling, the article focuses on the overall system architecture, the improvements made to the system based on preliminary tests and simulations, and finally the results of a field campaign. The Barbados experimental flight campaign confirmed the capacity of the system to map clouds and to collect relevant data in dynamic environment, and highlighted areas for improvement.

Simulation is very helpful for the development of cyber-physical systems, as it enables testing functionalities and their integration without full hardware deployment. For complex systems, such as fleets of heterogeneous robots, multiple simulators dedicated to particular physical processes must be interconnected, so as to build a wholesome simulation and test the overall system. A key property to ensure is that the overall simulation is repeatable. We propose a lightweight distributed architecture for time management, allowing to easily deploy complex simulations while strictly ensuring repeatability. A formal model of the architecture is provided, along with a proof of progress. An open source implementation, with a binding to the robotic ROS framework is made available.

Field multi-robot missions face numerous unavoidable disturbances, such as delays in executing tasks and intermittent communications. Coping with such disturbances requires to endow the robots with high-level decision skills. We present a distributed decision architecture based first on a hybrid planner that can manage decentralized repairs with partial communication, and secondly on a distributed execution algorithm that efficiently propagates delays. This architecture has been successfully experimented on the field for the achievement of surveillance missions involving eight (8) real autonomous aerial and ground robots.

We present an approach to plan trajectories for a fleet of fixed-wing UAVs to observe a wildfire evolving over time. Realistic models of the terrain, of the fire propagation process, and of the UAVs are exploited, together with a model of the wind. The approach tailors a generic Variable Neighborhood Search method to these models and associated constraints. Simulation results show ability to plan observation trajectories for a small fleet of UAVs, and to update the plans when new information on the fire are incorporated in the fire model.

Three-dimensional point clouds acquired with lidars are an important source of data for the classification of outdoor environments by autonomous terrestrial robots. We propose a two-layer classification model. The first layer consists of a Gaussian mixture model. This model is determined in a training step in an unsupervised manner and classified into a large set of classes. The second layer consists of a grouping of these classes. This grouping is determined by an expert during the training step and leads to a smaller set of classes that are interpretable in a considered target task. Because the first layer relies on unsupervised learning, manual labeling of data is not required. Supervision is necessary only for the second layer and in this case is assisted by the classes provided by the first layer. The evaluation is done for two data sets acquired with different lidars and possessing different characteristics. It is done quantitatively using one of the data sets and qualitatively using another. The system design follows a standard learning procedure with training, validation, and test steps. The operation follows a standard classification pipeline. The system is simple with no requirement of preprocessing or postprocessing stages.

A novel method for visual place recognition is introduced and evaluated, demonstrating robustness to perceptual aliasing and observation noise. This is achieved by increasing discrimination through a more structured representation of visual observations. Estimation of observation likelihoods are based on graph kernel formulations, utilizing both the structural and visual information encoded in covisibility graphs. The proposed probabilistic model is able to circumvent the typically difficult and expensive posterior normalization procedure by exploiting the information available in visual observations. Furthermore, the place recognition complexity is independent of the size of the map. Results show improvements over the state-of-theart on a diverse set of both public datasets and novel experiments, highlighting the benefit of the approach.

The lack of depth information in camera images has triggered much work on their use for localization and mapping in robotics. In particular, specific landmark parametrizations that isolate the unknown depth in one variable, and that allows to handle the associated large uncertainties have been proposed. Recently, an innovative parametrization (Parallax Angle) has shown to outperform the others in the context of a Bundle Adjustment approach. This paper investigates the way to exploit this parametrization in an incremental graph-based SLAM approach, in a robotics context in which motions measures can be incorporated in the overall estimation. It presents the factors required to initialize landmarks and manage their observations. Simulation results show that the proposed algorithms are able to incrementally incorporate observations, and a discussion analyzes how the incremental updates on ISAM2 are affected by these new factors.

We study the problem of monitoring the evolution of atmospheric variables within low-altitude cumulus clouds with a fleet of Unmanned Aerial Vehicles (UAVs). To tackle this challenge, two main problems can be identified: i) creating on-line maps of the relevant variables, based on sparse local measurements; ii) designing a planning algorithm which exploits the obtained map to generate trajectories that optimize the adaptive data sampling process, minimizing the uncertainty in the map, while steering the vehicles within the air flows to generate energetic-efficient flights. Our approach is based on Gaussian Processes (GP) for the mapping, combined with a stochastic optimization scheme for the trajectories generation. The system is tested in simulations carried out using a realistic three-dimensional current field. Results for a single UAV as well as for a fleet of multiple UAVs, sharing information to cooperatively achieve the mission, are provided.

The complexity of robotics comes from the tight interactions between hardware, complex softwares, and environments. While real world experience is the only way to assess the efficiency and robustness of a robotics system, simulations help to pave the way to actual experiments. But an overall robotics system requires simulations at a level of realism which no holistic simulator can provide, given the wide spectrum of disciplines and physical processes involved. This paper presents a way to integrate various simulators, in a distributed, scalable and repeatable way, to benefit from their different advantages and get the best fitted and accurate simulation for a given robotics system. It depicts how the MORSE open-source robotics simulator is adapted to comply with the High Level Architecture standard, thus allowing the reuse of numerous dedicated realistic simulators. Two examples of the integration of simulators are provided.

Environment models are the primary matter to autonomous decisions for mobile robots, and also to cooperation within teams of robots that operate in the same environment. The decisions to take within a robot or a robot team relate to motions, perceptions and communications: various types of environment models are therefore required to evaluate and plan these actions. While the literature abounds with approaches to environment modeling using data perceived by the robots, very few work tackle the problem of managing such models within a team of robots. Managing environment models implies first defining the proper data structures and associated mechanisms that allow both their efficient update and use by the decisional processes that require them, and second ensuring the models consistency as the robots evolve. This article presents the definition of a framework dedicated to the managing of environment models within a robot team. It establishes the principles that govern the framework design, and illustrates them throughout some examples.

With the growing demand for deployment of robots in real scenarios, robustness in the perception capabilities for navigation lies at the forefront of research interest, as this forms the backbone of robotic autonomy. Existing place recognition approaches traditionally follow the feature-based bag-of-words paradigm in order to cut down on the richness of information in images. As structural information is typically ignored, such methods suffer from perceptual aliasing and reduced recall, due to the ambiguity of observations. In a bid to boost the robustness of appearance-based place recognition, we consider the world as a continuous constellation of visual words, while keeping track of their covisibility in a graph structure. Locations are queried based on their appearance, and modelled by their corresponding cluster of landmarks from the global covisibility graph, which retains important relational information about landmarks. Complexity is reduced by comparing locations by their graphs of visual words in a simplified manner. Test results show increased recall performance and robustness to noisy observations, compared to state-of-the-art methods.

This article depicts an algorithm which matches the output of a Lidar with an initial terrain model to estimate the absolute pose of a robot. Initial models do not perfectly fit the reality and the acquired data set can contain an unknown, and potentially large, proportion of outliers. We present an interval based algorithm that copes with such conditions, by matching the Lidar data with the terrain model in a robust manner. Experimental validations using different terrain model are reported to illustrate the performance of the method.

Besides precise and dense geometric information, some LiDARs also provide intensity information and multiple echoes, information that can advantageously be exploited to enhance the performance of the purely geometric classification approaches. This information indeed depends on the physical nature of the perceived surfaces, and is not strongly impacted by the scene illumination - contrary to visual information. This article investigates how such information can augment the precision of a point cloud classifier. It presents an empirical evaluation of a low cost LiDAR, introduces features related to the intensity and multiple echoes and their use in a hierarchical classification scheme. Results on varied outdoor scenes are depicted, and show that more precise class identification can be achieved using the intensity and multiple echoes than when using only geometric features.

Bayesian filtering often relies on a reduced system state relating to robot internal variables only. The exogenous variables and their effects on the measurement process are then encompassed within a global observation noise model. Even if Bayes filters proved to be robust to such approximations, special care has to be taken to handle some of these exogenous effects, usually by introducing complex observation distributions or rejection rules. No matter how complex these models are, they often fail in dealing with contextual incidence which can hardly be explicitly encoded. This article shows how contextual information can be introduced within the Bayesian filtering framework by coupling a filter with classification and regression probabilistic models. The classification model provides an efficient context-dependent measurement selection mechanism and is specifically trained with respect to the filter estimation performance. This first component is enhanced by the introduction of context-dependent observation noise provided by the regression model. The performance of this is approach is evaluated and compared with other methods in the context of altitude estimation for a UAV.

The fusion of multi-sensor information for state estimation is a well studied problem in robotics. However, the classical methods may fail to take into account the measurements validity, therefore ruining the benefits of sensor redundancy. This work addresses this problem by learning context-dependent knowledge about sensor reliability. This knowledge is later used as a decision rule in the fusion task in order to dynamically select the most appropriate subset of sensors. For this purpose we use the Mixture of Experts framework. In our application, each expert is a Kalman filter fed by a subset of sensors, and a gating network serves as a mediator between individual filters, basing its decision on sensor inputs and contextual information to reason about the operation context. The performance of this model is evaluated for altitude estimation of a UAV.

In order to diminish the influence of pose choice during appearance-based mapping, a more natural representation of location models is established using covisibility graphs. As the robot moves through the environment, visual landmarks are detected, and connected if seen as covisible. The introduction of a novel generative model allows relevant subgraphs of the covisibility map to be compared to a given query without needing to normalize over all previously seen locations. The use of probabilistic methods provides a unified framework to incorporate sensor error, perceptual aliasing, decision thresholds, and multiple location matches. The system is evaluated and compared with other state-of-the-art methods.

This paper presents a framework to organize the various processes that endow a robot with autonomy. The main objectives are to allow the achievement of a variety of missions without an explicit writing of control schemes by the developer, and the possibility to augment the robot capacities without any major rewriting. The proposed architecture relies on a partition of the decisional layer in separate resources, each one managed by a specific agent. The mechanisms that guarantee the good use of each resource and manage the network of agent into a coherent system are depicted, and illustrated in the case of an autonomous navigation mission.

This paper presents a technique for the calibration of multi-beam laser scanners. The technique is based on an optimization process, which gives precise estimation of calibration parameters starting from an initial estimate. The optimization process is based on the comparison of scan data with the ground truth environment. Detailed account of the optimization process and suitability analysis of optimization objective function is described, and results are provided to show the efficacy of calibration technique.

We present a control strategy that permits a fixed wing UAV to visually track a ground target. A pursuit-evasion strategy aims at optimizing the visibility of the pursuer UAV after having predicted the best action for the evading target. This is achieved by using two iterative methods that optimise various mission based criteria: obstacle avoidance and visibility maximisation for the UAV and stealthy motions for the target.

A large-scale mapping approach is combined with multiple robots events to achieve cooperative mapping. The mapping approach used is based on hierarchical SLAM -global level and local maps-, which is generalized for the multi-robot case. In particular, the consequences of multi-robot loop closing events (common landmarks detection and relative pose measurement between robots) are analyzed and managed at a global level. We present simulation results for each of these events using aerial and ground robots, and experimental results obtained with ground robots.

We present a fast line segment tracker which does not require any knowledge about the motion of the camera nor the structure of the observed scene. It runs on 320 times 240 pixel images at 30 Hz. We adapted the RAPiD tracker with a new way of handling multiple line hypotheses to deal with the simple model of a single line segment. We discuss the difficulty of using a chi2-test as merging criterion and also present a new approach to overcome it. Furthermore, instead of making assumptions about the camera motion, a constant velocity motion model to predict the line segment position in the following frame is used. We explain how to deal with the instability of the endpoint extraction in this motion model to avoid unintentional motion along the line. Finally, we present results on real world indoor and urban outdoor image sequences.

In the context of stereovision SLAM, we propose a way to enrich the landmark models. Vision-based SLAM approaches usually rely on interest points associated to a point in the Cartesian space: by adjoining oriented planar patches (if they are present in the environment), we augment the landmark description with an oriented frame. Thanks to this additional information, the robot pose is fully observable with the perception of a single landmark, and the knowledge of the patches orientation helps the matching of landmarks. The paper depicts the chosen landmark model, the way to extract and match them, and presents some SLAM results obtained with such landmarks.

We present a simple control strategy to visually track a ground target with a fixed wing UAV. The approach uses a lateral guidance law that aims at reaching a given target view angle, which is determined on the basis of simple geometric considerations. Simulations compare our approach with previous contributions in terms of target visibility ratio, and results obtained during flight tests are presented.

This paper presents a method to incorporate 3D line segments in vision based SLAM. A landmark initialization method that relies on the Plucker coordinates to represent a 3D line is introduced: a Gaussian sum approximates the feature initial state and is updated as new observations are gathered by the camera. Once initialized, the landmarks state is estimated along an EKF-based SLAM approach: constraints associated with the Plucker representation are considered during the update step of the Kalman filter. The whole SLAM algorithm is validated in simulation runs and results obtained with real data are presented.

In military missions in hostile environments involving teams of UAVs flying in formation, it is important to get the maximum benefits of the auto-protection systems of each aircraft to enhance the global security and efficiency of the team. One way to achieve this is to select a proper configuration for the formation. In this paper, we present an approach to autonomously adapt the configuration of a formation and we focus on its evaluation within a realistic framework where each UAV is simulated independently and communicate through a network.

This paper presents a software component, the plan database, which provides the needed services to define plans, execute them and more importantly adapt them during execution. This plan database handles fully dynamic plans (insertion and removal of tasks), defines task transformation operators and provides tools for safe concurrent execution and modification of plans. These features are essential in multirobot and human-robot contexts, where tasks need to be easily passed between systems and plan adaptation helps coping with the unpredictability inherent to systems where multiple agent make decisions.

This paper presents an approach to detect safe landing areas for a flying robot, on the basis of a sequence of monocular images. The approach does not require precise position and attitude sensors: it exploits the relations between 2D image homographies and 3D planes. The combination of a robust homography estimation and of an adaptive thresholding of correlation scores between registered images yields the update of a stochastic grid, that exhibits the horizontal planar areas perceived. This grid allows the integration of data gathered at various altitudes. Results are presented

International audience

This video presents a control system that autonomously coordinates and supervises a fleet of heterogeneous UAVs to achieve complex observation missions. The system is composed of a central planning and monitoring station, and rely on a UAV decisional architecture that is designed to fit various levels of autonomy. The architecture and its adaptation to various kinds of UAVs is described, focusing on the role of the UAVs supervisors. A demonstration that illustrates a fire detection, confirmation and monitoring scenario, completed with a mapping task is shown. It involves three UAVs : two helicopters and a blimp, each of them being endowed with a supervisor linked to the central station. During the demonstration, the UAVs achieve various observation tasks coordinated thanks to their supervisor.

In the classic EKF SLAM framework, the vision based approaches raise the problem of landmark initialization. As a single observation is not enough to compute a gaussian state estimate, a special initialisation algorithm must be developed. This video presents our approach with a step by step simulation, then a more complex simulation is proposed. The last part gathers results obtained on real data.

The particular geometry of panoramic cameras defines complex epipolar lines equations. In this paper, we present a way to warp images from a panoramic stereovision bench, so that the epipolar lines become parallel straight lines, thus allowing the use of an optimized fast pixel correlation based stereovision algorithm. The paper first introduces the geometric characterization of panoramic camera composed of parabolic and spherical mirrors, that computes both the intrinsic parameters of the system (mirror surfaces and intrinsic camera parameters) and the errors alignment between the mirrors. Then, it presents the warping equations that allow to generate rectified images. Calibration and stereovision results are presented.

This paper presents a decisional architecture and the associated algorithms for multi-UAV (unmanned aerial vehicle) systems. The architecture enables different schemes of decision distribution in the system, depending on the available decision making capabilities of the UAVs and on the operational constraints related to the tasks to achieve. The paper mainly focuses on the deliberative layer of the UAVs: we detail a planning scheme where a symbolic planner relies on refinement tools that exploit UAVs and environment models. Integration effort related to decisional features is highlighted, and preliminary simulation results are provided.

This article presents a bearing only 3D SLAM algorithm which has the same complexity and optimality as the usual extended Kalman filter used in classical SLAM. We especially focus on the landmark initialization process, which relies on visual point features tracked in the sequence of acquired images: a probabilistic approach to estimate their parameters is presented. This induces a particular structure of the filter architecture, in which are memorized a set of past robot poses. Simulations are made to compare the influence of some parameters required by our approach, and results with an indoor robot and an airship are presented.

This paper presents an approach to autonomously monitor the behavior of a robot endowed with several navigation and locomotion modes, adapted to the terrain to traverse. The mode selection process is done in two steps: the best suited mode is firstly selected on the basis of initial information or a qualitative map built on-line by the robot. Then, the motions of the robot are monitored by various processes that update mode transition probabilities in a Markov system. The paper focuses on this latter selection process: the overall approach is depicted, and preliminary experimental results are presented.

This paper deals with the task allocation problem in multi-robot systems. We propose a completely distributed architecture, where robots dynamically allocate their tasks while they are building their plans. We first focus on the problem of simple "goto" tasks allocation: our approach involves an incremental task allocation algorithm based on the Contract-Net protocol. We introduce a parameter called equity coefficient in order to equilibrate the workload between the different robots and to control the triggering of the auction process. Then, we address the problem raised by temporal constraints between tasks by dynamically specifying temporary hierarchies among the tasks. Tests run in simulation quantify the benefits of our improvements.

This paper presents an approach to build high resolution digital elevation maps from a sequence of unregistered low altitude stereovision image pairs. The approach first uses a visual motion estimation algorithm that determines the 3D motions of the cameras between consecutive acquisitions, on the basis of visually detected and matched environment features. An extended Kalman filter then estimates both the 6 position parameters and the 3D positions of the memorized features as images are acquired. Details are given on the filter implementation and on the estimation of the uncertainties on the feature observations and motion estimations. Experimental results show that the precision of the method enables to build spatially consistent very large maps.

This paper presents a new hybrid path planning approach for autonomous robots navigating in natural unknown environments. The main purpose is to consider perception planning and path planning in a single unified process, so that the most relevant perception can be performed considering the current goal of the robot. We consider this approach as a way to fill the gap between navigation tasks and exploration tasks. In a first part, we introduce the models and notions used in this approach. We then give some algorithmic details, and finally present results in simulation.

This article presents an approach to improve and monitor the behavior of a skid-steering rover on rough terrains. An adaptive locomotion control generates speeds references to avoid slipping situations. An enhanced odometry provides a better estimation of the distance travelled. A probabilistic classification procedure provides an evaluation of the locomotion efficiency on-line, with a detection of locomotion faults. Results obtained with a Marsokhod rover are presented throughout the paper.

Describes the design and implementation of an algorithm for improving the performance of stereo vision in environments presenting repetitive patterns or regions with relatively weak texture. The proposed algorithm makes use of the common assumption that the disparities corresponding to continuous surfaces in the world vary smoothly; we use this assumption to alleviate the correspondence problem for pixels that cannot be reliably matched by the stereo algorithm. Our approach can be described as a reliability based filtering of the disparity image followed by a recursive propagation step. It can be applied to the output of almost any "standard" stereo algorithm with minimal modifications, and is computationally efficient.

In this paper, we present an approach to qualitative rover localization with panoramic images. The approach relies on the possibility to efficiently and robustly compute the resemblance between panoramic images, indexing them by histograms of local appearances. A database of image indexes is dynamically built during rover motions: when the rover re-perceives an already crossed area, it matches the current image with the stored ones (place recognition), and thus gets a qualitative estimate of its position. Experimental results on a 400 images database illustrates the effectiveness of the algorithms.

This paper presents the current status of the development of our autonomous blimp project. Details are given on the hardware setup, which is currently almost operational. Some first experimental results on terrain mapping with low altitude stereo imagery are presented and discussed. The approach involves the integration of several algorithms: stereovision, interest point matching, motion estimation and digital elevation map building.

This paper presents an algorithm that matches interest points detected on a pair of grey level images taken from arbitrary points of view. First matching hypotheses are generated using a similarity measure of the interest points. Hypotheses are confirmed using local groups of interest paints: group matches are based on a measure defined on an affine transformation estimate and on a correlation coefficient computed on the intensity of the interest points. Once a reliable match has been determined for a given interest point and the corresponding local group, new group matches are found by propagating the estimated affine transformation. The algorithm has been widely tested under various image transformations: it provides dense matches and is very robust to outliers, i.e. interest points generated by noise or present in only one image because of occlusions or non overlap.

This article presents an algorithm that determines safe motions for an articulated rover on rough terrains. It relies on the evaluation of a set of elementary trajectories on a digital elevation map built as the rover moves. The algorithm relies on the explicit computation of geometric constraints on the rover chassis. It has been integrated within a continuously running navigation loop on board the robot Lama, and tested under various terrain conditions.

Presents a method that estimates robot displacements in outdoor unstructured terrain. It computes the displacements on the basis of associations of 3D points sets produced by consecutive stereovision frames, the associations being determined by tracking pixels from one image frame to the other. The paper details the various steps of the algorithms, and first experimental results are presented: they show that the algorithm is able to estimate the 6 parameters of the robot position with a relative error smaller than about 5%, processing several hundreds of images over several tens of meters.

The paper presents an approach to reactive navigation in cross-country terrains. The approach relies on a particular probabilistic obstacle detection procedure, that describes the area perceived by a pair of stereo cameras as a set of polygonal cells. To generate the motion commands on the basis of this terrain description, we present some improvements and adaptations to the classical potential fields technique. Results on real stereo data illustrate our contribution throughout the paper, and simulated long range traverses are discussed.

We present an approach to refine the pose estimate of an outdoor mobile robot evolving on flat terrains cluttered with obstacles. We propose an algorithm to extract relevant obstacle contour lines on the basis of stereo-vision data. The algorithm is very robust with respect to the uncertainties on the data, and do not require a very fine and precise obstacle extraction procedure. We explain how the contour lines are compared from one image to another to refine the pose estimate provided by the robot internal sensors.

We consider the case of a robot that must find ways in an initially unknown and often complex cross-country environment. Our approach relies on a particular model of the environment built from 3D data: the environment is represented by a hierarchical polygonal structure, in which the probabilities of terrain classes are updated as the robot moves. We describe how to define traversability costs within such a structure, taking into account the label probabilities, and explain how a shortest path search can produce a sequence of polygonal cells that must be crossed.

We are interested in inferring the sources of various types of sonar features typically observed by a mobile robot. After a brief discussion of terrestrial sonar sensing, we develop a set of operators that associates arc-shaped features extracted from sonar scans with real world primitives. Our classification scheme is probabilistic and is based on empirical data: the confidence of the association hypotheses produced by the operators is evaluated statistically. Some of our experimental results suggest that methods based on models of perfect sonar sensors may not be completely consistent with observed data. The management and merging of a collection of hypotheses concerning various sonar features allows the system to produce a coherent and mutually-compatible set of inferences for the entire observed environment.

This paper presents the approach, algorithms and processes we developed to perform cross-country autonomous navigation. After a presentation of the teleprogramming context, we introduce an adaptive navigation approach, well suited for the characteristics of complex natural environments. The main perception, motion planning and decisional processes required by the robot during navigation are briefly presented. An on board control architecture that manages all these processes is then described, and first results of an experiment currently developed at LAAS are discussed.>

The authors present a method for selecting viewpoints and sensing tasks to confirm by a multisensory perception machine, an identification hypothesis previously generated. The determination relies on the use of a compiled knowledge base that links object and sensor models, and defines a priori the best sensing tasks to be performed. The method is fully detailed, and its use under dynamic constraints due to a real environment is explained, along with a control strategy to activate the search. Result of experiments integrating other modules for environment modeling, path planning and execution control, and object recognition are presented.>