A "divide and conquer" strategy based on situations to achieve reactive collision avoidance in troublesome scenarios

Abstract

Abstract-This paper addresses the reactive collision avoidance for robots that move in arduous environments (i.e. very dense, complex and cluttered). To achieve this goal, the technique simplifies the difficulty of the navigation by a divide and conquer strategy, which is based on identifying navigational situations and applying the corresponding motion laws. The state of the art in reactive navigation still presents classic limitations such as the trap situations due to U-shape obstacles, the difficulty to achieve motion among very close obstacles, to obtain oscillation-free and stable motion, to move over directions far from the goal direction or towards the obstacles, or to tune the heuristic or internal parameters. This paper presents a method that overcomes all these limitations. As a result, navigation with this method is successfully achieved in scenarios where existing techniques present a high degree of difficulty to navigate. Outstanding navigation results are reported using a wheelchair vehicle and a discussion and comparison with existing techniques is provided. I. INTRODUCTIONCurrently, applications such as emergency response, transportation, deminning, exploration, help care or ludic robots yield the motion control to autonomous navigation systems. The mobile devices, provided with these capabilities, greatly improve their performance and the security of the potential occupants or transported objects. Thus in mobile robotics, we are carrying out a vast effort to improve the robustness of the basic skills of the mobility task, since it is clear that they have a large impact over the whole performance of the system. The navigation systems work over the vehicles with full autonomy or supervising the human operation. A good example is a robotic wheelchair, which must support both functionalities: complete autonomy (in high degrees of human disability) or supervising the operator orders (in lower degrees of disability). Furthermore, this vehicle requires motion in unstructured, unknown and dynamic scenarios with a high density of obstacles (e.g. moving among chairs, door traversal, humans moving around etc). The navigation performance is crucial for the normal development of this task (human transportation). This paper focuses on the motion generation aspect, which is the particular relevance here because collisions put in risk the safety of the occupant and sudden motions lead to an uncomfortable behavior. We present in this paper a technique able to achieve this objective.Classically, the motion planning problem is solved by computing a geometrical trajectory avoiding known obstacles (see [15] for a review of techniques). However, the general methods for motion planning are not applicable if the environment is dynamic with a priori unknown behavior, or if it is gradually discovered. Moreover, when both the environment model and

Sensor-based robot motion generation in unknown, dynamic and troublesome scenarios

“…In some cases, the reactive navigation module was replaced by another ND method adapted to work on platforms with a non-circular geometry and kinematic and dynamic constraints [57]. The substitution of these modules within the architecture was straightforward.…”

An Approach to Distributed Component-Based Software for Robotics

“…This is a real application we have usually in operation on the robots we have at the laboratory. The system implements a secure navigation system based on the ND+ algorithm for obstacle avoidance [Minguez et al (2004)]. It has been implemented attending to ].…”

Integration of planning and reactive obstacle avoidance in autonomous sensor-based navigation

“…The ND+ method [20] is an improvement of the ND method [17]. The design is based on defining a set of situations to represent the navigational problem, and how to act in each of them (actions).…”

“…The ND+ method improves the previous ND method with new navigational situations and a new design of the motion laws (to have motion continuity in the most common transitions between situations). Another advantage of the ND+ method is that is very efficient and it can be used when required without imposing a significant time penalty [20].…”

“…With regard to the planners, these systems use efficient numerical techniques on grids that are executed in real time [4], [25]. Another key issue is the obstacle avoidance method, where some systems use the potential field methods [13], those based on intermediate sets of commands [6], [24], [12], or those based on high-level information [17], [20].…”