This paper addresses the problem of path planning in environments in which some of the obstacles can change their positions. It uses the popular PRM method for navigating a robot through an environment. One of the key features of PRM is that it moves the major part of the calculations involved in the path planning process to the preprocessing phase. After that, paths can be extracted very quickly (in a query phase) usually without any noticeable delay. While very successful in many applications, doing most of the work in a preprocessing phase restricts the environment to be static i.e. obstacles are not allowed to change their configurations after the preprocessing phase. In this paper we describe and evaluate an algorithm based on PRM that does allow obstacles to change their configuration after preprocessing while still allowing for a quick query phase.

This paper addresses the problem of maneuvering an object by pushing it through an environment with obstacles. Instead of only pushing the object through open areas, we also allow it to use compliance, e.g., allowing it to slide along obstacle boundaries. Using compliance has a number of advantages: it extends the number of situations in which a manipulation plan can be found, it allows for simpler (i.e., less complicated) paths in many cases, and it often helps solving narrow-passage problems. Here, we present an approach based on rapidly-exploring random trees. Our approach yields paths through the open space, but also exploits the power of compliance.

This paper addresses the problem of maneuvering an object by pushing it through an environment with obstacles. Instead of only pushing the object through open spaces, we also allow it to use compliance, e.g. allowing it to slide along obstacle boundaries. The advantage of using compliance is twofold: compliance does not only extend the number of situations in which a push plan can be found, it also allows for simpler (i.e. less complicated) paths in many cases. Here, we present an approach based on the rapidly-exploring random tree (RRT) algorithm that, besides paths through the open space, exploits the power of compliance

In this paper we introduce a method based on the probabilistic roadmap (PRM) planner to construct robust roadmaps for motion planning in changing environments. PRM's are usually aimed at static environments. In reality though, many environments are not static, but contain moving obstacles as well. Often the motion of these obstacles is not unconstrained, but is restricted to some confined area, e.g. a door that can be open or closed or a chair which is bounded to a room. We exploit this observation by assuming that a moving obstacle has a predefined set of potential placements. We present a variant of PRM that is robust against placement changes of obstacles. Our method creates a roadmap that is guaranteed to contain a path for any feasible query when time goes to infinity, i.e. the method is probabilistically complete. Our implementation shows that after a roadmap is created in the preprocessing phase, queries can be solved instantaneously, thus allowing for on-the-fly replanning to anticipate changes in the environment.

We consider the path planning problem for a robot that pushes a disk shaped object in an environment among obstacles. Instead of only allowing the object to move through the free space, we also allow the object to slide along the boundaries of the environment using compliance, extending the possibilities for the robot to find a push path. We present an exact algorithm that, given a path for the object consisting of k sections, preprocesses the environment consisting of n non-intersecting line segments in O(n/sup 2/ log n) and reports a push path in O(kn log n) time or reports failure if no path exists. Under the weak assumption of low obstacle density, the query time is reduced to O((k + n) log n).

Over the last decade, the probabilistic road map method (PRM) has become one of the dominant motion planning techniques. Due to its random nature, the resulting paths tend to be much longer than the optimal path despite the development of numerous smoothing techniques. Also, the path length varies a lot every time the algorithm is executed. We present a new technique that results in higher quality (shorter) paths with much less variation between the executions. The technique is based on adding useful cycles to the roadmap graph.

Moving a camera through a (virtual or real) environment is a complicated task. Often a user is given direct control of the camera. Such direct control is difficult for inexperienced users and results in rather ugly camera motions that easily lead to motion sickness. In this paper we describe a new technique for automatic generation of camera motion using motion planning techniques from robotics. We focus here on motion in virtual environments. In our approach the user simply specifies a required goal position (and orientation) using e.g. a map, and the system automatically computes a smooth, collision free motion from the current position and orientation to the required position (and orientation) that can be used by the camera. As preprocessing, the approach uses the probabilistic roadmap method to compute a roadmap through the environment. When a motion is required, a path is obtained from the roadmap which is then improved by various smoothing techniques to satisfy constraints from cinematography.