Objective-based stochastic framework for manipulation planning

Steven M. La Valle, Seth A. Hutchinson

Research output: Contribution to conferencePaper

Abstract

We consider the problem of determining robot manipulation plans when sensing and control uncertainties are specified as conditional probability densities. Traditional approaches are usually based on worst-case error analysis in a methodology known as preimage backchaining. We have developed a general framework for determining sensor-based robot plans by blending ideas from stochastic optimal control and dynamic game theory with traditional preimage backchaining concepts. We argue that the consideration of a precise loss (or performance) functional is crucial to determining and evaluating manipulation plans in a probabilistic setting. We consequently introduce a stochastic, performance preimage that generalizes previous preimage notions. We also present some optimal strategies for planar manipulation tasks that were computer by a dynamic programming-based algorithm.

Original languageEnglish (US)
Pages1772-1779
Number of pages8
StatePublished - Dec 1 1994
EventProceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems. Part 3 (of 3) - Munich, Ger
Duration: Sep 12 1994Sep 16 1994

Other

OtherProceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems. Part 3 (of 3)
CityMunich, Ger
Period9/12/949/16/94

Fingerprint

Robots
Planning
Game theory
Dynamic programming
Error analysis
Sensors
Uncertainty

ASJC Scopus subject areas

  • Engineering(all)

Cite this

La Valle, S. M., & Hutchinson, S. A. (1994). Objective-based stochastic framework for manipulation planning. 1772-1779. Paper presented at Proceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems. Part 3 (of 3), Munich, Ger, .

Objective-based stochastic framework for manipulation planning. / La Valle, Steven M.; Hutchinson, Seth A.

1994. 1772-1779 Paper presented at Proceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems. Part 3 (of 3), Munich, Ger, .

Research output: Contribution to conferencePaper

La Valle, SM & Hutchinson, SA 1994, 'Objective-based stochastic framework for manipulation planning', Paper presented at Proceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems. Part 3 (of 3), Munich, Ger, 9/12/94 - 9/16/94 pp. 1772-1779.
La Valle SM, Hutchinson SA. Objective-based stochastic framework for manipulation planning. 1994. Paper presented at Proceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems. Part 3 (of 3), Munich, Ger, .
La Valle, Steven M. ; Hutchinson, Seth A. / Objective-based stochastic framework for manipulation planning. Paper presented at Proceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems. Part 3 (of 3), Munich, Ger, .8 p.
@conference{2076f840e7f94ea8995863f7c0ee5841,
title = "Objective-based stochastic framework for manipulation planning",
abstract = "We consider the problem of determining robot manipulation plans when sensing and control uncertainties are specified as conditional probability densities. Traditional approaches are usually based on worst-case error analysis in a methodology known as preimage backchaining. We have developed a general framework for determining sensor-based robot plans by blending ideas from stochastic optimal control and dynamic game theory with traditional preimage backchaining concepts. We argue that the consideration of a precise loss (or performance) functional is crucial to determining and evaluating manipulation plans in a probabilistic setting. We consequently introduce a stochastic, performance preimage that generalizes previous preimage notions. We also present some optimal strategies for planar manipulation tasks that were computer by a dynamic programming-based algorithm.",
author = "{La Valle}, {Steven M.} and Hutchinson, {Seth A.}",
year = "1994",
month = "12",
day = "1",
language = "English (US)",
pages = "1772--1779",
note = "Proceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems. Part 3 (of 3) ; Conference date: 12-09-1994 Through 16-09-1994",

}

TY - CONF

T1 - Objective-based stochastic framework for manipulation planning

AU - La Valle, Steven M.

AU - Hutchinson, Seth A.

PY - 1994/12/1

Y1 - 1994/12/1

N2 - We consider the problem of determining robot manipulation plans when sensing and control uncertainties are specified as conditional probability densities. Traditional approaches are usually based on worst-case error analysis in a methodology known as preimage backchaining. We have developed a general framework for determining sensor-based robot plans by blending ideas from stochastic optimal control and dynamic game theory with traditional preimage backchaining concepts. We argue that the consideration of a precise loss (or performance) functional is crucial to determining and evaluating manipulation plans in a probabilistic setting. We consequently introduce a stochastic, performance preimage that generalizes previous preimage notions. We also present some optimal strategies for planar manipulation tasks that were computer by a dynamic programming-based algorithm.

AB - We consider the problem of determining robot manipulation plans when sensing and control uncertainties are specified as conditional probability densities. Traditional approaches are usually based on worst-case error analysis in a methodology known as preimage backchaining. We have developed a general framework for determining sensor-based robot plans by blending ideas from stochastic optimal control and dynamic game theory with traditional preimage backchaining concepts. We argue that the consideration of a precise loss (or performance) functional is crucial to determining and evaluating manipulation plans in a probabilistic setting. We consequently introduce a stochastic, performance preimage that generalizes previous preimage notions. We also present some optimal strategies for planar manipulation tasks that were computer by a dynamic programming-based algorithm.

UR - http://www.scopus.com/inward/record.url?scp=0028731892&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028731892&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0028731892

SP - 1772

EP - 1779

ER -