J. Kosecka and R. Bajcsy, "Discrete Event Systems for Autonomous Mobile Agents", Intelligent Robotic Systems, `93, Zakopane, 1993 (also to appear in Robotics and Autonomous Systems, 12(3) march 94.  Home/Search   Document Details and Download   Summary   Related Articles   Check

..... Priority based arbitration: which is a subsumptive style, priority based arbitration mechanism, where behaviors with higher priorities are allowed to suppress the output of behaviors with lower priorities. State based arbitration: which is based on the Discrete Event Systems (DES) formalism [7], and is suitable for behvior sequencing. Command fusion mechanisms combine recommendations from multiple behaviors to form a control action that represents their consensus. This approach provides for a coordination scheme that allows all behaviors to simultaneously contribute to the control of ....

J. Kosecka and R. Bajcsy, "Discrete Event Systems for autonomous mobile agents," in Proc. Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....computational expense of grid based maps. robotics as the gain and loss of constraints due to obstacles are fundamentally discrete events which cause abrupt changes in the permissible dynamics of the robot. Discrete event methods have been considered for mobile robots previously by Kosecka et al. [9, 10, 8] who defined the control commands as discrete states (e.g. moving, steer away, path following) How ever, these models are rather arbitrary without a mathematical basis for the state definitions. Instead, we propose a constraint based method for defining the discrete states. As the mobile robot ....

J. Kosecka and R. Bajcsy. Discrete event systems for autonomous mobile agents. Journal of Robotics and Autonomous Systems, 12(B14):187-198, 1994.

....robot change in a discrete fashion as the robot is constrained by obstacles. Traditional methods for mobile robot control neglect the discrete aspect of the system. Discrete event methods have been considered for mo bile robots previously, but only by a few researchers. In particular, Kosecka [Kosecka and Bajcsy, 1994; Kosecka et al. 1995; Kosecka, 1996] considered discrete event modelling of navigation of mobile robots with vari ous tasks defined as discrete states (e.g. moving, steer away, path following) However, these models are rather arbitrary without a mathematical basis for the state definitions. ....

....A rigourous discrete event modelling technique for nonholonomic mobile robots is proposed. This technique recognises that the dynamics of the mobile robot change in a discrete fashion as the robot becomes constrained by obstacles. Unlike previous discrete event models for mobile robots [Kosecka and Bajcsy, 1994; Kosecka et al. 1995; Kosecka, 1996] the model proposed here is rigourous, using the presence of obstacles to define the discrete state. The discrete state also depends upon the dynamics of the robot so that an accurate model of the obstacle is provided for control purposes. A dynamic model ....

J. Kosecka and R. Ba- jcsy. Discrete event systems for autonomous mobile agents. Journal of Robotics and Autonomous Systems, 12(B14) :187-198, 1994.

....completely formal, and therefore is suitable for formulating, and ultimately proving, correctness properties for monitored systems. Plans for ongoing and future work include the following issues: 1. Draw closer parallels with the concept of controllable systems in discrete event control theory [20, 6, 14]. 2. Explore realizations of execution monitors different than that presented in Section 4.2 [28] 3. Investigate techniques for proving correctness properties of various monitors and monitored systems. 4. Investigate the concept of execution monitoring for non terminating Golog programs [11] ....

J. Kosecka and R. Bajcsy. Discrete Event Systems for Autonomous Mobile Agents. Robotics and Autonomous Systems, volume 12, p. 187--198, 1994.

....and behavior is also influenced by its mental state whether it is happy or sad, excited or not etc. Since Bouncy s behavior repertoire is rather limited it was possible to control the activation of its behaviors using a straight forward action selection mechanism known as Discrete Event Systems [2], which is based on the finite state automaton (FSA) formalism. Figure 3 depicts the finite state machine that is used to describe Bouncy s action selection mechanism. It is seen that there is one state corresponding to each Behavior. State transitions are caused by events denoted tired, rested ....

Jana Koseck'a and Ruzena Bajcsy. Discrete Event Systems for Autonomous Mobile Agents. . Proceedings Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....state whether it is happy or sad, excited or not etc. Bouncy s liveliness behaviors are mutually exclusive, i.e. only one can be active at a time. Thus it is possible to control the activation of its behaviors using a straight forward action selection mechanism known as Discrete Event Systems [2], which is based on the finite state automaton (FSA) formalism. Figure 6 depicts the finite state machine that is used to describe Bouncy s action selection mechanism. It is seen that there is one state corresponding to each Behavior. State transitions are caused by events denoted tired, rested ....

Jana Koseck'a and Ruzena Bajcsy. Discrete Event Systems for Autonomous Mobile Agents. . Proceedings Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....GCSR toolkit to experiment the expressiveness and limitations of GCSR and the refinement theory through real or complex existing real time systems. We are considering examples from avionics such as the TCASS II system proposed by Nancy Leveson, robotics such as the Autonomous Mobile Agents [KB94] or computer animation. In addition to being real time applications, such examples are usually complex and thus require an incremental development of the system specification. Appendix A Selected Proofs A.1 Proofs of Hoare Ordering Extension A.1.1 Proof of Theorem 4.1 Proof: Assume that P ....

Jana Kosecka and Ruzena Bajcsy. Discrete event systems for autonomous mobile agents. Robotics and Autonomous Systems, 12:187--198, 1994.

....state whether it is happy or sad, excited or not etc. Bouncy s liveliness behaviors are mutually exclusive, i.e. only one can be active at a time. Thus it is possible to control the activation of its behaviors using a straight forward action selection mechanism known as Discrete Event Systems [2], which is based on the finite state automaton (FSA) formalism. Figure 6 depicts the finite state machine that is used to describe Bouncy s action selection mechanism. It is seen that there is one state corresponding to each Behavior. State transitions are caused by events denoted tired, rested ....

Jana Koseck'a and Ruzena Bajcsy. Discrete Event Systems for Autonomous Mobile Agents. . Proceedings Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....under varying conditions. The subsumption architecture [2] is a priority based arbitration mechanism, where behaviors with higher priorities are allowed to subsume the output of behaviors with lower priorities. State based arbitration mechanisms include the Discrete Event Systems (DES) formalism [5] and a similar method called temporal sequencing [1] Finally, in winner take all mechanisms action selection results from the interaction of a set of distributed behaviors that compete until one wins and takes control of the system [6] Command fusion mechanisms combine recommendations from ....

Jana Koseck'a and Ruzena Bajcsy. Discrete Event Systems for Autonomous Mobile Agents. . Proceedings Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....to be relevant. The Subsumption architecture [5] is a priority based arbitration mechanism, where behaviors with higher priorities are allowed to suppress the output of behaviors with lower priorities. State based arbitration mechanisms include the Discrete Event Systems (DES) formalism [12], a similar method called temporal sequencing [3] and action selection based on Bayesian decision theory [13] Finally, in winner take all mechanisms action selection results from the interaction of a set of distributed behaviors that compete until one wins and takes control of the system, such ....

Jana Koseck'a and Ruzena Bajcsy. Discrete Event Systems for Autonomous Mobile Agents. Proceedings Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....investigated by many researchers, where its advantages and disadvantages were explored and demonstrated. 7 State Based arbitration 7. 1 Discrete Event Systems [Koseck a 1993] In this framework the interaction of behaviors are formalized by the DES formalism in a modular and hierarchical fashion [23]. Behavior selection is done using state transition, where upon detection of a certain event a shift is made to a new state and thus a new behavior. Using this formalism, systems are modeled in terms of finite state automata (FSA) where states correspond to execution of actions behaviors and ....

....similar to the discipline of, e.g. classical control theory. However, we realize that in order to allow the formalism to capture the essence of a system one should make certain approximations and simplifications. In the dynamical systems approach, the fuzzy approach [57] and the DES formalism [23] it is seen that stability, formal proofs of task achievement and controllability guarantees, respectively, can be given when the approach is based on a formal framework. Multiple objective behavior coordination mechanisms provide formal tools for characterizing action selection in terms of ....

Jana Koseck'a and Ruzena Bajcsy. Discrete Event Systems for Autonomous Mobile Agents. . Proceedings Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....robot change in a discrete fashion as the robot is constrained by obstacles. Traditional methods for mobile robot control neglect the discrete aspect of the system. Discrete event methods have been considered for mobile robots previously, but only by a few researchers. In particular, Kosecka [ Kosecka and Bajcsy, 1994; Kosecka et al. 1995; Kosecka, 1996 ] considered discrete event modelling of navigation of mobile robots with various tasks defined as discrete states (e.g. moving, steer away, path following) However, these models are rather arbitrary without a mathematical basis for the state definitions. ....

....A rigourous discrete event modelling technique for nonholonomic mobile robots is proposed. This technique recognises that the dynamics of the mobile robot change in a discrete fashion as the robot becomes constrained by obstacles. Unlike previous discrete event models for mobile robots [ Kosecka and Bajcsy, 1994; Kosecka et al. 1995; Kosecka, 1996 ] the model proposed here is rigourous, using the presence of obstacles to define the discrete state. The discrete state also depends upon the dynamics of the robot so that an accurate model of the obstacle is provided for control purposes. A dynamic model ....

J. Kosecka and R. Bajcsy. Discrete event systems for autonomous mobile agents. Journal of Robotics and Autonomous Systems, 12(B14):187--198, 1994.

....priority based, state based and winner take all. The subsumption architecture[3] is a priority based arbitration mechanism, where behaviors with higher priorities are allowed to subsume the output of behaviors with lower priority. State based arbitration mechanisms include Discrete Event Systems[8], a similar method called temporal sequencing[2] and action selection based on Bayesian decision theory[9] The advantage of the DES and the temporal sequencing approaches is that they are based on the finite state machines formalism and thus system observeability and controllability can be shown ....

Jana Koseck'a and Ruzena Bajcsy. Discrete Event Systems for Autonomous Mobile Agents. . Proceedings Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....The subsumption architecture [3] promoted by Brooks, is a priority based arbitration mechanism, where behaviors with higher priorities are allowed to subsume the output of behaviors with lower priority. State based arbitration mechanisms include the Discrete Event Systems (DES) formalism [9], a similar method called temporal sequencing described in [2] and action selection based on Bayesian decision theory [10] The advantage of the DES and the temporal sequencing approaches is that they are based on the finite state machines formalism and thus system observeability and ....

Jana Koseck'a and Ruzena Bajcsy. Discrete Event Systems for Autonomous Mobile Agents. . Proceedings Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....the process of validating that all conditions are complimentary literals. The TR subroutines designed for robot control using SPOTT are typically concise and thus can be easily checked for completeness (see Section III C. 3) TR programs also bear some resemblance to discrete event systems [36], 13] in their formulation and graphical representation. One difference is that inputs are implicit in the TR conditions while TR arcs explicitly represent just outputs (rather than both input and output, as in the arcs of Petri nets [37] In addition, TR conditions (i.e. nodes) monitor ....

J. Kosecka and R. Bajcsy, "Discrete event systems for autonomous mobile agents," Robotics and Autonomous Systems, vol. 12, pp. 187--198, 1994.

....give a geometrical impression of the accuracy of the detection. 3. 6 Obstacle Detector The visual obstacle detects obstacles by inverse perspective transformation from the image plane to the ground plane of a pair of stereo images, followed by an absolute subtraction of the two transformed images [9, 3]. The obstacle detection is obtained by thresholding the difference image, and subsequently filtering it with a morphological erosion. The remaining pixels in the image are labeled as obstacles, and the pixels filtered out, are labeled as free space. The advantage of this method is that the ....

Jana Kosecka and Ruzen Bajcsy. Discrete event systems for autonomous mobile agents. Proceedings Intelligent Robotic Systems '93 Zakopane, pages 21--31, July 1993.

....well the interactions between participating components and at the same time supports modular and hierarchical design principles. This work falls into the category of hybrid approaches and offers some systematic means for the design of the mediator process. Our previous experience in [KB93b; KB94] justified the feasibility of the DES modeling. However, as the system becomes more complicated and the coupling between components more task dependent, the process of specification of the desired behavior becomes more tedious. The domain of cooperating autonomous mobile agents offers a large ....

Jana Kosecka and Ruzena Bajcsy. Discrete event systems for autonomous mobile agents. Robotics and Autonomous Systems, 1994.

....transitions between the states are created by external events (depending on the task) or by the successful completion or failure of the strategy. Each fundamental process is modeled as a DES system as described by its associated automaton G = Q; Sigma; ffi; q 0 ; Qm ) in the fashion defined in [19]. Events are classified into two categories: controllable events Sigma c (the ones which may be prevented from occurring or forced to occur) and uncontrollable events Sigma u (those which may not be prevented from occurring) Controllable events identify basic strategies (which run continuously ....

J. Kosecka and R. Bajcsy. Discrete event systems for autonomous mobile agents. In Workshop on Intelligent Robot Control, Zakopane, July 1993.

No context found.

J. Kosecka and R. Bajcsy, "Discrete Event Systems for Autonomous Mobile Agents", Intelligent Robotic Systems, `93, Zakopane, 1993 (also to appear in Robotics and Autonomous Systems, 12(3) march 94.

No context found.

J. Kosecka and R. Bajcsy, "Discrete event systems for autonomous mobile agents," Robotics and Autonomous Systems, vol. 12, no. 3-4, pp. 187--198, 1994.

No context found.

J. Kosecka and R. Bajcsy, "Discrete event systems for autonomous mobile agents," Robotics and Autonomous Systems, vol. 12, no. 3-4, pp. 187--198, 1994.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC