FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W89169610> ?p ?o ?g. }

• W89169610 abstract "This position paper is concerned with the implications of limited computational resources and uncertainty on sensing and planning in robotic systems. To address the computational complexity of sensor interpretation .lnd planning processes, we redefine their principal role. Following Agrc and Chapman's plan-as-communication approach, sensing and planning are treated as computational processes that provide information to an execution archileclure and thus improve the overall performance of the system. We argue that robots must be able to Hade off the quality of this information against its computational costs. Anytime algorithms, whose quality of results improves gradually as computation time increases, provide useful performance componenl~ for tidle-critical sensing and planning in robotic systems. The paper describes some uf our current research directions and open problems. The role of sensing and planning Sensing1 and planning are computational processes aimed at helping an autonomous robot (or an agent) act intelligently in the world . The goal of sensing and planning is to improve the performance of an agent by providing certain types of information. A similar idea, termed plan-as-communication, was introduced by Agre and Chapman (1990). Agre and Chapman contrasted this approach with the more traditional approaches that they term planas-program. The plan-as-program view has several disadvantages reiated to its computational complexity, inability to handle imprecise sensory data and uncertalllty. It also requires that plans be too detailed and it fails to relate the plan to the concrete situation . We share Agre and Chapman's view of traditional planning. Our approach is similar to their plan-ascommunication idea, but we offer an additional important step, Since senslllg and planning provide information to the • Author's address: University of Massachusetts, Computer Science Department. LGRC, Box 3461 0, Amherst, MA 01003-4610. Email: shlomo@cs.umass.edu. Phone: (413) 545-4189. URL: http://anytime,cs , umass.edu/~shlomo. IThe term Sensing refers to the computational task of perceptual data interpretation rather then the mechanical process of sensing. l4l execution architecture, their execution should be managed based on the value of this information. Moreover, since the quality of the information degrades over time, an agent should trade off the quality of the information against the computation time needed to produce it. At the heart of any robotic system lies an execution architecture that interacts with the physical world by performing certain external actions. The goal of sensing and planning is to provide the execution architecture with information that can improve the selection of actions. We distinguish between the actions selected by the planner to construct the plan and the actions selected by the execution architecture which are the actual actions performed by the robot. The two processes may be completely different and may sometimes disagree. The performance improvement that is achieved by the sensing and planning processes is reflected by the system reaching more of its goals in less time. The sensing and planning modules of a robot should be designed to support a particular execution architecture. Unfortunately, much of the work on sensing and planning within the AI community has been carried out with respect to either tri vial or ill-specified execution architectures. In reality, planning must be accompanied by an execution architecture that can translate the plan into individual external actions, monitor their execution , recognize various types of failures , recover from simple failures, and otherwise rely on re-planning. Some recent attempts to design planning systems have recognized the central role of execution architecture. For example, the Entropy Reduction Engine (ERE) (Bresina and Drummond 1990) is deSigned to integrate pianning and reaction for control of NASA's autonomous rover. The ERE architecture includes three components , reactor, projector and reducer, that have the capability to operate independently and perform their assigned task. Planning performed by the projector is provided as advice to the reactor in order to improve tts performance. Similarly, the reductor provides the projector with search control advice. The robotics community has a long tradition of integrating planning with reaction . The typical approach has been to introduce a planner on top of a reacti ve execution architecture. For example, SROMA (Xiaodong and Bekey 1988) is an adaptive scheduler for mechanical assembly From AAAI Technical Report SS-96-04. Compilation copyright © 1996, AAAI (www.aaai.org). All rights reserved." @default.
• W89169610 created "2016-06-24" @default.
• W89169610 creator A5027224308 @default.
• W89169610 date "2002-01-01" @default.
• W89169610 modified "2023-09-27" @default.
• W89169610 title "Resource-Bounded Sensing and Planning in Robotic Systems" @default.
• W89169610 cites W1496432202 @default.
• W89169610 cites W1525110644 @default.
• W89169610 cites W1589640093 @default.
• W89169610 cites W184814775 @default.
• W89169610 cites W1976129448 @default.
• W89169610 cites W2000005166 @default.
• W89169610 cites W2001863269 @default.
• W89169610 cites W2036190613 @default.
• W89169610 cites W2058038825 @default.
• W89169610 cites W2068965944 @default.
• W89169610 cites W2079927714 @default.
• W89169610 cites W2127668611 @default.
• W89169610 cites W2131857895 @default.
• W89169610 cites W2161117570 @default.
• W89169610 cites W2162412417 @default.
• W89169610 cites W2166996723 @default.
• W89169610 cites W2170370992 @default.
• W89169610 cites W2250180543 @default.
• W89169610 cites W4251996 @default.
• W89169610 hasPublicationYear "2002" @default.
• W89169610 type Work @default.
• W89169610 sameAs 89169610 @default.
• W89169610 citedByCount "0" @default.
• W89169610 crossrefType "journal-article" @default.
• W89169610 hasAuthorship W89169610A5027224308 @default.
• W89169610 hasConcept C111472728 @default.
• W89169610 hasConcept C11413529 @default.
• W89169610 hasConcept C127413603 @default.
• W89169610 hasConcept C138885662 @default.
• W89169610 hasConcept C144559511 @default.
• W89169610 hasConcept C154945302 @default.
• W89169610 hasConcept C166957645 @default.
• W89169610 hasConcept C179799912 @default.
• W89169610 hasConcept C206345919 @default.
• W89169610 hasConcept C2776505523 @default.
• W89169610 hasConcept C2779530757 @default.
• W89169610 hasConcept C31258907 @default.
• W89169610 hasConcept C38652104 @default.
• W89169610 hasConcept C41008148 @default.
• W89169610 hasConcept C42475967 @default.
• W89169610 hasConcept C66024118 @default.
• W89169610 hasConcept C90509273 @default.
• W89169610 hasConcept C95457728 @default.
• W89169610 hasConceptScore W89169610C111472728 @default.
• W89169610 hasConceptScore W89169610C11413529 @default.
• W89169610 hasConceptScore W89169610C127413603 @default.
• W89169610 hasConceptScore W89169610C138885662 @default.
• W89169610 hasConceptScore W89169610C144559511 @default.
• W89169610 hasConceptScore W89169610C154945302 @default.
• W89169610 hasConceptScore W89169610C166957645 @default.
• W89169610 hasConceptScore W89169610C179799912 @default.
• W89169610 hasConceptScore W89169610C206345919 @default.
• W89169610 hasConceptScore W89169610C2776505523 @default.
• W89169610 hasConceptScore W89169610C2779530757 @default.
• W89169610 hasConceptScore W89169610C31258907 @default.
• W89169610 hasConceptScore W89169610C38652104 @default.
• W89169610 hasConceptScore W89169610C41008148 @default.
• W89169610 hasConceptScore W89169610C42475967 @default.
• W89169610 hasConceptScore W89169610C66024118 @default.
• W89169610 hasConceptScore W89169610C90509273 @default.
• W89169610 hasConceptScore W89169610C95457728 @default.
• W89169610 hasLocation W891696101 @default.
• W89169610 hasOpenAccess W89169610 @default.
• W89169610 hasPrimaryLocation W891696101 @default.
• W89169610 hasRelatedWork W1574208191 @default.
• W89169610 hasRelatedWork W1980021494 @default.
• W89169610 hasRelatedWork W2033873806 @default.
• W89169610 hasRelatedWork W2044372685 @default.
• W89169610 hasRelatedWork W2159056570 @default.
• W89169610 hasRelatedWork W2162301465 @default.
• W89169610 hasRelatedWork W2767575257 @default.
• W89169610 hasRelatedWork W2784226822 @default.
• W89169610 hasRelatedWork W2792006456 @default.
• W89169610 hasRelatedWork W2793611303 @default.
• W89169610 hasRelatedWork W2943178670 @default.
• W89169610 hasRelatedWork W2954160944 @default.
• W89169610 hasRelatedWork W2984075506 @default.
• W89169610 hasRelatedWork W3004021781 @default.
• W89169610 hasRelatedWork W3015177126 @default.
• W89169610 hasRelatedWork W3037956663 @default.
• W89169610 hasRelatedWork W3135954914 @default.
• W89169610 hasRelatedWork W3206746502 @default.
• W89169610 hasRelatedWork W32855506 @default.
• W89169610 hasRelatedWork W1527704360 @default.
• W89169610 isParatext "false" @default.
• W89169610 isRetracted "false" @default.
• W89169610 magId "89169610" @default.
• W89169610 workType "article" @default.