TangentBug: A Range-Sensor-Based Navigation Algorithm

Abstract

The Bug family algorithms navigate a 2-DOF mobile robot in a completely unknown environment using sensors. TangentBug is a new algorithm in this family, specifically designed for using a range sensor. TangentBug uses the range data to compute a locally shortest path, based on a novel structure termed the local tangent graph (LTG). The robot uses the LTG for choosing the locally optimal direction while moving toward the target, and for making local shortcuts and testing a leaving condition while moving along an obstacle boundary. The transition between these two modes of motion is governed by a globally convergent criterion, which is based on the distance of the robot from the target. We analyze the properties of TangentBug, and present simulation results that show that Tangent-Bug consistently performs better than the classical Bug algorithms. The simulation results also show that TangentBug produces paths that in simple environments approach the globally optimal path, as the sensor's maximal detection-range increases. The algorithm can be readily implemented on a mobile robot, and we discuss one such implementation.

Algorithms for collision-free navigation of mobile robots in complex cluttered environments: a survey

“…• A method of probabilistically convergent on-line navigation involves randomly choosing tangents to travel down (see, e.g., [237]), or by use of the deterministic TangentBug algorithm (see, e.g., [119]). Tangent events can be detected from a ray-based sensor model (see, e.g., [246]) or by processing data from a camera sensor (see, e.g., [113]).…”

Simple concurrency for robotics with the Roboscoop framework

“…Exemplary classes in the subcluster include WANDER BEHAVIOR and TANGENT BUG BEHAVIOR, an implementation of the tangent bug algorithm [18]. …”

Boundary Following and Globally Convergent Path Planning Using Instant Goals

“…It can be deduced that the likelihood of obtaining a better global performance will increase by choosing a locally preferable direction at each Start. In [18], the initial boundary following direction was chosen based on the orientation of the boundary at the hit point (similar to H 1 i in this paper) and expected that following the direction would take the robot closer to the goal. However, this approach utilizes only partial information, that is, the information of the hit point.…”

“…In order to guarantee convergence to the goal, the transition between the two modes is governed by a criterion to ensure that the distance to the goal decreases monotonously. To improve navigation performance such as shorter paths, DistBug [2] and TangentBug [18] algorithms adapt the tangent graph to obtain the locally shortest path, define several different transition conditions for switching between the two motion modes, and choose the locally optimal direction for obstacle following. Similar to the Bug algorithms, the subgoal selection algorithm [19] generates a sequence of subgoals near the polygonal vertices, until the obstacle does not block the line-of-sight to the final goal.…”