2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2021
DOI: 10.1109/iros51168.2021.9636346
|View full text |Cite
|
Sign up to set email alerts
|

TrajectoTree: Trajectory Optimization Meets Tree Search for Planning Multi-contact Dexterous Manipulation

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
11
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
6
2

Relationship

0
8

Authors

Journals

citations
Cited by 20 publications
(11 citation statements)
references
References 24 publications
0
11
0
Order By: Relevance
“…With this in mind, Sundaralingam and Hermans ( 2018 ) presented a planner for reorientation of the object through finger gaiting and in-grasp manipulation alternately. Similarly, Chen C. et al ( 2021a ) proposed TrajectoTree, a method based on contact-implicit trajectory optimization (CITO). Unlike the optimization method, the concept of motion primitives is also accepted widely (Chen C. et al, 2021b ; Yoneda et al, 2021 ).…”
Section: Dexterous Manipulation For Multi-fingered Robotic Hand Based...mentioning
confidence: 99%
See 1 more Smart Citation
“…With this in mind, Sundaralingam and Hermans ( 2018 ) presented a planner for reorientation of the object through finger gaiting and in-grasp manipulation alternately. Similarly, Chen C. et al ( 2021a ) proposed TrajectoTree, a method based on contact-implicit trajectory optimization (CITO). Unlike the optimization method, the concept of motion primitives is also accepted widely (Chen C. et al, 2021b ; Yoneda et al, 2021 ).…”
Section: Dexterous Manipulation For Multi-fingered Robotic Hand Based...mentioning
confidence: 99%
“…However, only under certain assumptions can these approaches work, such as assuming that the shape and mass of the object are known and the contacts remain during the manipulation process (Sundaralingam and Hermans, 2017 ). Also, some approaches can only be applied to planar objects (Chen C. et al, 2021a ). At the same time, most of these methods are only tested in simulation.…”
Section: Dexterous Manipulation For Multi-fingered Robotic Hand Based...mentioning
confidence: 99%
“…On the other hand, the contact mode sequence is inherently discrete and the optimizer faces a fundamentally discrete choice at each time, which is difficult to optimize whether modeled using continuous constraints or integer variables. To that end, recent approaches use graph search-based methods [20][21][22] or rapidly-exploring random tree [23] to plan contact switches and generate a seed for the subsequent trajectory optimization. However, these local methods are greedy and do not offer a fall-back in case the trajectory optimization does not succeed using the discrete contact sequence.…”
Section: P Wmentioning
confidence: 99%
“…So we used a method that is fairly common to generate smooth trajectories for manipulation planning. Also we intend to show the benefits of interleaving graph search and contact implicit trajectory optimization over the common choice of using them in sequence [20][21][22][23]. To do so, we compare our method with a sequential combination of bi-directional RRT and direct collocation [23].…”
Section: Insat: Interleaved Search and Trajectory Optimizationmentioning
confidence: 99%
“…Recently, an algorithm that augments Contact-Implicit Trajectory Optimization (CITO) with tree search was proposed in [16] to incorporate domain-specific knowledge for robot manipulation. It uses Depth First Search (DFS) to find a sequence of kinematically feasible contact modes that have a stable grasp and then constrain the CITO problem with the found contact sequence.…”
Section: Introductionmentioning
confidence: 99%