Saffiotti, A.: Some notes on the integration of planning and reactivity in autonomous mobile robots, Proceedings of the AAAI Spring Symposium on Foundations of Planning, 1993.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....using example scenarios. We conclude in Section 6. 2 Related Work There is a vast literature on the integration of planning and reactivity in autonomous mobile robots, e.g. on the traditional sense plan act architectures as well as behaviour based control and variations thereof [Gat, 1992; Saffiotti, 1993] Several of the issues related to that are discussed in the following sections. However, a detailed discussion of the previous work on this topic goes beyond the scope of this paper. In this section, we focus on literature on the realisation of reasoningbased execution monitoring and on ....

A. Saffiotti. Some notes on the integration of planning and reactivity in autonomous mobile robots. In Proc. of the AAAI Spring Symposium on Foundations of Planning, pages 122--126, Stanford, CA, US, 1993.

.... shown by Saffiotti [11] a fuzzy based controller has the advantage that the intuitive nature of collision free navigation can be easily modelled using linguistic terminology (e.g. next right instead of 89 25 o on precise radius) 14] This implies that the low level behaviour 3 Saffiotti [17] gives an interesting glance at the planning vs. reactivity debate information exploration commands for module cartographic generation map path planning planner module map planner location path navigator sensing control actions command list topologic map navigator (task ....

....on high precision, metric approaches and demand very reliable and accurate sensor devices as well as great computational power. In addition, these methods have difficulties during navigation in complex and dynamically changing environments where unknown obstacles appear on an a priori planned path [17, 14]. The conventional path planning approaches for mobile robots may be divided into two categories. One is global path planning based on a priori complete information about the environment. The other is the local path planning based on sensor information in uncertain environment where size, shape ....

A. Saffiotti, "Some Notes on the Integration of Planning and Reactivity in Autonomous Mobile Robots," in Procs. of the AAAI Spring Symposium on Foundations of Automatic Planning, (Stanford, CA), pp. 122 -- 126, 1993.

.... mobili hanno evidenziato che essi devono essere in grado di raggiungere degli obiettivi di alto livello tenendo conto dei cambiamenti dinamici a cui e soggetto l ambiente che li circonda, devono cio e integrare opportunamente la componente reattiva con il sistema di ragionamento e pianificazione [4, 12]. In questo lavoro si propone di usare un sistema per la rappresentazione della conoscenza di tipo generale basato sul formalismo delle logiche descrittive. Ci o e possibile poich e, partendo dalla formalizzazione di un sistema dinamico in termini di una teoria di una logica dinamica ....

....start a x, ovvero un piano in cui lo stato finale soddisfa G. 3 Pianificazione CLASSIC in Tino La teoria esposta nella sezione precedente e stata implementata, tenendo conto delle restrizioni imposte dal linguaggio CLASSIC , nel robot mobile Tino . L architettura di Tino e a due livelli [4, 12]. Il livello di controllo e implementato da un controllore fuzzy [13] che si occupa di interpretare i dati sensoriali e di eseguire alcuni comportamenti basilari tra cui quello che permette di evitare gli ostacoli. Il livello di pianificazione invece e costituito da una base di conoscenza CLASSIC ....

A. Saffiotti. Some notes on the integration of planning and reactivity in autonomous mobile robot. Foundations of Automatic Planning.

....the RealWorld Navigator: that the navigation component can be fixed across application domains. Instead, a common approach is to allow higher level components to activate, deactivate or parameterize navigation processes. Recent examples include ATLANTIS [ Gat, 1992 ] SSS [ Connell, 1992 ] and [ Saffiotti, 1993 ] A further alternative is to compile beforehand a reactive structure that will execute a plan at run time (again, navigation is neither a fixed component nor a necessary part of these structures) Examples include [ Kaelbling and Rosenschein, 1989 ] Schoppers, 1987 ] and [ Nilsson, 1994 ] ....

Alessandro Saffiotti. Some notes on the integration of planning and reactivity in autonomous mobile robots. In AAAI Spring Symposium on Automated Planning, 1993.

....actions are imprecise and may fail. Classical planning approaches to the control of mobile robots have been criticized for not being able to adequately cope with uncertainty, imprecision and Currently at: Iridia, Universit e Libre de Bruxelles, Belgium, asaffio ulb.ac. be execution errors [4, 7, 5, 12]. To overcome these limitations, some authors have proposed architectures centered around a complex controller, or reactive planner (e.g. 4, 7] These reactive architectures provide immediate response to unpredicted environmental situations by giving up the idea of reasoning about future ....

A. Saffiotti. Some notes on the integration of planning and reactivity in autonomous mobile robots. In Procs. of the AAAI Spring Symposium on Foundations of Automatic Planning, pages 122--126, Menlo Park, California, 1993. AAAI Press.

....landmark recognition. One of the components of a control structure is an artifact representing an object associated with a behavior to be performed, e.g. to cross a door or to follow a wall. Execution of control structures is based on anchoring the object of the action to a physical object [ Saffiotti et al. 1993 ] For example, consider a Follow behavior whose associated object is a wall. In this case, the artifact is a representation of an abstract wall, located at an expected location (e.g. determined from the map) When the action is executed, perception is used to try to match a perceived wall like ....

....detected by the sonars) to this artifact. If the match is successful e.g. the position and orientation are within certain tolerance ranges then the artifact is anchored to the perceived object and its properties are updated accordingly. In our case, the 2 We have also made experiments [Saffiotti et al. 1993] using more sophisticated AI planners, like SIPE [Wilkins, 1988] and real time decision systems, like PRS [Georgeff and Ingrand, 1989] a) b) Figure 3: Corridor following before anchoring (a) and after anchoring (b) position of the artifact is modified to reflect the position of the ....

[Article contains additional citation context not shown here]

Saffiotti, A. 1993. Some notes on the integration of planning and reactivity in autonomous mobile robots. In Procs. of the AAAI Spring Symposium on Foundations of Automatic Planning, Stanford, CA. 122--126.

....can be manipulated by a procedure. In order to perform an action, the agent needs to anchor its internal descriptions of the objects of an action to physical objects in the environment. These problems, and the first two in particular, have been the subject of intensive study in recent years (e.g. [1, 3, 18, 8, 11, 4, 13, 15]) In this note, we focus on the third problem. In pick up(A) we propose to represent the object of the action, A , by a description of the class of physical objects that are suitable for our action. That is, we represent this object by the set of properties that make an object the right one ....

A. Saffiotti. Some notes on the integration of planning and reactivity in autonomous mobile robots. In Procs. of the AAAI Spring Symposium on Foundations of Automatic Planning, pages 122--126, Stanford, CA, 1993.

....a plan to exit building E could consist in three successive corridors to follow, one control point in the entrance hall close to the door, and the exit door itself. The context of applicability of each control structure is used to decide when each control structure becomes relevant. see [ Saffiotti et al. 1993; Saffiotti, 1993 ] for more on this issue) We believe that having based our architecture on fuzzy logic results in improved robustness (e.g. more tolerance to sensor noise and knowledge imprecision) while granting a better understanding of the underlying mechanisms. Finally, many current ....

....E could consist in three successive corridors to follow, one control point in the entrance hall close to the door, and the exit door itself. The context of applicability of each control structure is used to decide when each control structure becomes relevant. see [ Saffiotti et al. 1993; Saffiotti, 1993 ] for more on this issue) We believe that having based our architecture on fuzzy logic results in improved robustness (e.g. more tolerance to sensor noise and knowledge imprecision) while granting a better understanding of the underlying mechanisms. Finally, many current approaches to robot ....

Saffiotti, A. 1993. Some notes on the integration of planning and reactivity in autonomous mobile robots. In Procs. of the AAAI Spring Symposium on Foundations of Automatic Planning, Stanford, CA.

No context found.

Saffiotti, A.: Some notes on the integration of planning and reactivity in autonomous mobile robots, Proceedings of the AAAI Spring Symposium on Foundations of Planning, 1993.

No context found.

A. Saffiotti, "Some Notes on the Integration of Planning and Reactivity in Autonomous Mobile Robots", AAAI'93, pp. 122-126, 1993.

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