2026-09-20T01:55:55Zhttps://dspace.mit.edu/server/oai/request

oai:dspace.mit.edu:1721.1/1589462025-04-07T08:34:48Zcom_1721.1_7582com_1721.1_7581col_1721.1_131022

Tedrake, Russell L. Suh, Hyung Ju Terry Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science 2025-03-27T16:59:51Z 2025-03-27T16:59:51Z 2025-02 2025-03-04T17:26:00.786Z https://hdl.handle.net/1721.1/158946 Contact-rich manipulation has proved challenging due to the need to consider multiple combinatoric possibilities of making or breaking contact with the surrounding environment. As a result, existing methods have often resorted to combinatorial optimization that utilizes dynamics structure but considers all possibilities exhaustively, or compute-heavy and inefficient sampling methods that utilize blackbox optimization such as Reinforcement Learning (RL). In this thesis, I aim to show that by combining structured contact smoothing in conjunction with local gradient-based control and sampling-based motion planning, we can bypass the combinatorial explosion of contact modes while still leveraging structure and achieve highly efficient contact-rich manipulation. To achieve this capability, I first shed light on how RL abstracts contact modes and optimizes difficult landscapes by combining stochastic smoothing and zeroth-order optimization; yet, I show how following a similar stochastic strategy while using gradients suffers from several drawbacks such as empirical bias and high variance. To leverage structure in a more helpful manner, I propose a method for smoothing contact dynamics without relying on stochastic smoothing, bypassing these drawbacks. Using this smoothing scheme, I present a highly efficient and performant local control based on gradient-based trajectory optimization and model predictive control. Finally, I connect these local control capabilities with global sampling-based motion planners to achieve long-horizon global plans. The proposed method achieves contact-rich plans such as dexterous in-hand reorientation and whole-body manipulation much more efficiently than RL while being highly scalable compared to methods that explicitly reason about contact modes. These results achieve a reduction of contact-rich manipulation to kinodynamic motion planning, and exposes the true difficulty of contact-rich manipulation from combinatorial explosion in contact modes to combinatorial and highly non-local decisions over motion planning behaviors. Ph.D. Massachusetts Institute of Technology Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) Copyright retained by author(s) https://creativecommons.org/licenses/by-nc-nd/4.0/ Leveraging Structure for Efficient and Dexterous Contact-Rich Manipulation Thesis application/pdf Doctoral Doctor of Philosophy Publication unknown unknown <Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="d7fed93a-05d1-4653-8f58-965a20b0845d"> <Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type> <Title>Leveraging Structure for Efficient and Dexterous Contact-Rich Manipulation</Title> <PublishedIn> <Publication> </Publication> </PublishedIn> <PublicationDate>2025-02</PublicationDate> <Authors> <Author> <DisplayName>Suh, Hyung Ju Terry</DisplayName> <Affiliation> <OrgUnit> </OrgUnit> </Affiliation> </Author> </Authors> <Editors> </Editors> <Publishers> <Publisher> <DisplayName>Massachusetts Institute of Technology</DisplayName> <OrgUnit /> </Publisher> </Publishers> <License>https://creativecommons.org/licenses/by-nc-nd/4.0/</License> <Abstract>Contact-rich manipulation has proved challenging due to the need to consider multiple combinatoric possibilities of making or breaking contact with the surrounding environment. As a result, existing methods have often resorted to combinatorial optimization that utilizes dynamics structure but considers all possibilities exhaustively, or compute-heavy and inefficient sampling methods that utilize blackbox optimization such as Reinforcement Learning (RL). In this thesis, I aim to show that by combining structured contact smoothing in conjunction with local gradient-based control and sampling-based motion planning, we can bypass the combinatorial explosion of contact modes while still leveraging structure and achieve highly efficient contact-rich manipulation. To achieve this capability, I first shed light on how RL abstracts contact modes and optimizes difficult landscapes by combining stochastic smoothing and zeroth-order optimization; yet, I show how following a similar stochastic strategy while using gradients suffers from several drawbacks such as empirical bias and high variance. To leverage structure in a more helpful manner, I propose a method for smoothing contact dynamics without relying on stochastic smoothing, bypassing these drawbacks. Using this smoothing scheme, I present a highly efficient and performant local control based on gradient-based trajectory optimization and model predictive control. Finally, I connect these local control capabilities with global sampling-based motion planners to achieve long-horizon global plans. The proposed method achieves contact-rich plans such as dexterous in-hand reorientation and whole-body manipulation much more efficiently than RL while being highly scalable compared to methods that explicitly reason about contact modes. These results achieve a reduction of contact-rich manipulation to kinodynamic motion planning, and exposes the true difficulty of contact-rich manipulation from combinatorial explosion in contact modes to combinatorial and highly non-local decisions over motion planning behaviors.</Abstract> <Access xmlns="http://purl.org/coar/access_right" > </Access> </Publication>