Browsing by Author "Rybski, Paul E."
Now showing 1 - 3 of 3
- Results Per Page
- Sort Options
Item Performance Evaluation of a Multi-Robot Search & Retrieval System: Experiences with MinDART(2003-02-08) Rybski, Paul E.; Veeraraghavan, Harini; Lapoint, Monica; Gini, MariaThe costs of developing mobile robot teams can be reduced if they are designed to exploit swarm techniques. In this methodology many simple homogeneous units solve complex tasks through emergent behavior. The challenge lies in selecting an appropriate control strategy for the individual units. Complexity in design costs both money and time, therefore a control strategy should be just complex enough to perform the task successfully in a variety of environments, relative to some performance measure. To explore the effects of control strategies and environmental factors on performance, we have conducted two sets offoraging experiments using real robots (the {em Minnesota Distributed Autonomous Robotic Team}). The first set of experiments tested the efficacy of localization capabilities, in addition to the effects of team size and target distribution. The second set tested the efficacyof communication. We found that more complex control strategies do not necessarily improve task completion times, however they can reduce variance in performance measures. This can be valuable information for designers who need to assess the potential costs and benefits of increased complexity in design.Item Performance of a Distributed Robotic System Using Shared Communications Channels(2001-07-26) Rybski, Paul E.; Stoeter, Sascha A.; Gini, Maria; Hougen, DeanWe have designed and built a set of miniature robots, called Scouts. In addition, we have developed a distributed software system to control them. This paper addresses the fundamental choices we made in the design of the control software, describes experimental results in a surveillance task,and analyzes the factors that affect robot performance. Space and power limitations on the Scouts severely restrict the computational power of their on-board computer, which can only handle low-level control operations and data communication on an RF data link. We use a proxy-processing scheme, in which the robots are dependent on remote computers for their computing needs. While this allows the distributed robotics system to be autonomous, the fact that the robots behaviors are executed on remote computers introduces an additional complication -- the behavior controller has to receive sensor data and send commands to the robots using RF communication channels. Because the capacity of the communication system is limited, the robots must share bandwidth. We have developed a process management/scheduling system that is capable of handling high demand when controlling a group of robots. The resource allocation system dynamically assigns resources to each robot in an attempt to maximize the utilization of the available resources while still maintaining a priori behavior priorities. We present experimental results on a surveillance task in which multiple robots patrol an area and watch for motion. We discuss how the limited communication bandwidth affects robot performance in accomplishing the task and analyze how performance depends on the number of robots that share the bandwidth.Item Visual Servoing of a Miniature Robot Toward a Marked Target(2002-01-31) Rößler, Patrick; Stoeter, Sascha A.; Rybski, Paul E.The paper describes a system that servos a miniature robot toward a marked target based on visual information from the robot's own camera. This task is divided into two subtasks: recognition of the target and driving toward the target while updating the estimation of its position. This system deals with the problems introduced by noisy camera images and unstructured non-static backgrounds. A description of the control software is given and experimental results arediscussed.