Benjamín Tovar

Distance-Optimal Navigation in an Unknown Environment Without Sensing Distances

This paper considers what can be accomplished using a mobile robot that has limited sensing. For navigation and mapping, the robot has only one sensor, which tracks the directions of depth discontinuities. There are no coordinates, and the robot is given a motion primitive that allows it to move toward discontinuities. The robot is incapable of performing localization or measuring any distances or angles. Nevertheless, when dropped into an unknown planar environment, the robot builds a data structure, called the gap navigation tree, which enables it to navigate optimally in terms of Euclidean distance traveled. In a sense, the robot is able to learn the critical information contained in the classical shortest-path roadmap, although surprisingly it is unable to extract metric information. We prove these results for the case of a point robot placed into a simply connected, piecewise-analytic planar environment. The case of multiply connected environments is also addressed, in which it is shown that further sensing assumptions are needed. Due to the limited sensor given to the robot, globally optimal navigation is impossible; however, our approach achieves locally optimal (within a homotopy class) navigation, which is the best that is theoretically possible under this robot model.

Planning exploration strategies for simultaneous localization and mapping

In this paper, we present techniques that allow one or multiple mobile robots to efficiently explore and model their environment. While much existing research in the area of Simultaneous Localization and Mapping (SLAM) focuses on issues related to uncertainty in sensor data, our work focuses on the problem of planning optimal exploration strategies. We develop a utility function that measures the quality of proposed sensing locations, give a randomized algorithm for selecting an optimal next sensing location, and provide methods for extracting features from sensor data and merging these into an incrementally constructed map. We also provide an efficient algorithm driven by our utility function. This algorithm is able to explore several steps ahead without incurring too high a computational cost. We have compared that exploration strategy with a totally greedy algorithm that optimizes our utility function with a one-step-look ahead. The planning algorithms which have been developed operate using simple but flexible models of the robot sensors and actuator abilities. Techniques that allow implementation of these sensor models on top of the capabilities of actual sensors have been provided. All of the proposed algorithms have been implemented either on real robots (for the case of individual robots) or in simulation (for the case of multiple robots), and experimental results are given. c 2005 Elsevier B.V. All rights reserved.

Visibility-Based Pursuit-Evasion with Bounded Speed

In this paper we present a study on the visibility-based pursuit— evasion problem in which bounds on the speeds of the pursuer and evader are given. In this setting, the pursuer tries to find the evader inside a simply connected polygonal environment, and the evader in turn tries to avoid detection. The focus of the paper is to develop a characterization of the set of possible evader positions as a function of time (the reachable sets). This characterization is more complex than the unbounded-speed case, because it no longer depends only on the combinatorial changes in the visibility region of the pursuer. The characterization presented can be used as a filter to keep track of the possible evaders position as a pursuer moves in the environment, or it can be used to guide the construction of the pursuer search strategy. This search algorithm is at least as powerful as a complete algorithm for the unbounded-speed case, and with the knowledge of speed bounds, generates solutions for environments that were unsolvable previously. Given that numerical methods are needed to compute the reachable sets, we also present a conservative approximation which can be computed with a closed formula.

Optimal navigation and object finding without geometric maps or localization

In this paper we develop a dynamite data structure, useful for robot navigation in an unknown, simply connected planar environment. The guiding philosophy in this work is to avoid traditional problems such as complete map building and localization by constructing a minimal representation based entirely on critical events in online sensor measurements made by the robot. Furthermore, this representation provides a sensor-feedback motion strategy that guides the robot along an optimal trajectory between any two environment locations, and allows the search of static targets, even though there is no geometric map of the environment. We present algorithms for building the data structure in an unknown environment, and for using it to perform optimal navigation. We implemented these algorithms on a real mobile robot. Results are presented in which the robot builds the data structure online, and is able to use it without needing a global reference frame. Simulation results are shown to demonstrate how the robot is able to find interesting objects in the environment.

A reactive motion planner to maintain visibility of unpredictable targets

This paper deals with the problem of computing the motions of one or more robot observers in order to maintain visibility of one or several moving targets. The targets are assumed to move unpredictably, and the distribution of obstacles in the workspace is assumed to be known in advance. Our algorithm computes a motion strategy by maximizing the shortest distance to escape

A Sampling-Based Motion Planning Approach to Maintain Visibility of Unpredictable Targets

the shortest distance the target needs to move in order to escape the observers visibility region. Three main points are discussed: 1) the design and implementation of a reactive planner; 2) the integration and testing of such a planner in a robot system which includes perceptual and control capabilities; and 3) the design and simulation of a motion planner for the task of maintaining visibility of two targets using two mobile observers.

Sensor Beams, Obstacles, and Possible Paths

This paper deals with the surveillance problem of computing the motions of one or more robot observers in order to maintain visibility of one or several moving targets. The targets are assumed to move unpredictably, and the distribution of obstacles in the workspace is assumed to be known in advance. Our algorithm computes a motion strategy by maximizing the shortest distance to escape—the shortest distance the target must move to escape an observers visibility region. Since this optimization problem is intractable, we use randomized methods to generate candidate surveillance paths for the observers. We have implemented our algorithms, and we provide experimental results using real mobile robots for the single target case, and simulation results for the case of two targets-two observers.

Robot motion planning for map building

This paper introduces a problem in which an agent (robot, human, or animal) travels among obstacles and binary detection beams. The task is to determine the possible agent path based only on the binary sensor data. This is a basic filtering problem encountered in many settings, which may arise from physical sensor beams or virtual beams that are derived from other sensing modalities. Methods are given for three alternative representations: 1) the possible sequences of regions visited, 2) path descriptions up to homotopy class, and 3) numbers of times winding around obstacles. The solutions are adapted to the minimal sensing setting; therefore, precise estimation, distances, and coordinates are replaced by topological expressions. Applications include sensor-based forensics, assisted living, security, and environmental monitoring.

Locally-optimal navigation in multiply-connected environments without geometric maps

The goal of this work is to develop techniques that allow one or more robotic observers to operate with full autonomy while accomplishing the task of model building. The planning algorithm operates using certain simple but flexible models of the observer sensor and actuator abilities. We provide techniques that allow us to implement these sensor models on top of the capabilities of the actual (and off-the-shelf) sensors we have. It is worth keeping the following points in mind regarding our goals: 1) even with completely idealized sensing and mobility capabilities, the algorithmic task of model building is quite challenging; 2) computational techniques can be used to approximate and implement these idealized sensors on top of actual sensors; and 3) the quality and success of the generated plans depend significantly on the observer capabilities. The study of this dependency terms of high-level parameters describing the sensors is part of this work.

Information spaces for mobile robots

In this paper we present an algorithm to build a sensor-based, dynamic data structure useful for robot navigation in an unknown, multiply-connected planar environment. This data structure offers a robust framework for robot navigation, avoiding the need of a complete geometric map or explicit localization, by building a minimal representation based entirely on critical events in online sensor measurements made by the robot. There are two sensing requirements for the robot: it must detect when it is close to the walls, to perform wall-following reliably, and it must be able to detect discontinuities in depth information. It is also assumed that the robot is able to drop, detect and recover a marker. The navigation paths generated are optimal up to the homotopy class to which the paths belong, even though no distance information is measured.