Reliably arranging objects : a conformant planning approach to robot manipulation
Author(s)
Anders, Ariel(Ariel Sharone)
Download1102048229-MIT.pdf (19.82Mb)
Alternative title
Conformant planning approach to robot manipulation
Other Contributors
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science.
Advisor
Tomas Lozano-Perez and Leslie Pack Kaelbling.
Terms of use
Metadata
Show full item recordAbstract
A crucial challenge in robotics is achieving reliable results in spite of sensing and control uncertainty. In this work, we explore the conformant planning approach to reliable robot manipulation. In particular, we tackle the problem of pushing multiple planar objects simultaneously to achieve a specified arrangement without using external sensing. A conformant plan is a sequence of manipulation actions that reliably achieve a goal arrangement in spite of uncertainty in object pose and nondeterministic action outcomes, and without assuming the availability of additional observations. To find conformant plans, we explored two different approaches: Conformant planning by construction. This approach formalizes conformant planning as a belief-state planning problem. A belief state is the set of all possible states of the world, and the objective is to find a sequence of actions that will bring an initial belief state to a goal belief state. To do forward belief-state planning, we created a deterministic belief-state transition model from on-line physics-based simulations and supervised learning based on off-line physics simulations. Conformant planning through plan improvement. This approach takes a deterministic manipulation plan and augments it by adding fixtures (movable obstacles) to push parts up against. This method uses an optimization-based approach to determine the ideal fixture placement location. This thesis provides insight and develops approaches toward scalable methods for solving challenging planar manipulation problems with multiple objects or concave shape geometry. We show the success of these approaches based on planning times and robustness in real and simulated experiments.
Description
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. Thesis: Ph. D., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2019 Cataloged from student-submitted PDF version of thesis. Vita. Includes bibliographical references (pages 123-126).
Date issued
2019Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer SciencePublisher
Massachusetts Institute of Technology
Keywords
Electrical Engineering and Computer Science.