Sensor-based analysis of high-precision insertion tasks

Yamamoto, Y.; Yoneyama, Takehiro; Hashimoto, T.; Okubo, Tsuyoshi; Itoh, Takuzì

doi:10.1109/irds.2002.1044032

Cited by 3 publications

(4 citation statements)

References 13 publications

(15 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The existing controllers for tactile feedback manipulation often use control heuristics and need a model for the contact dynamics [3], [5], [11]. Recently, a number of methods combining force kinematics with compliant control and machine learning have been proposed to overcome the need of manual tuning [12], [13], [14], [15], [16].…”

Section: Related Workmentioning

confidence: 99%

“…The sense of touch is a key sensory modality for many robot manipulation tasks such as grasping [1] and precisioninsertion [2], [3], [4]. Tactile sensing is an instrumental modality of robotic manipulation, as it provides information that is not accessible via remote sensors such as cameras or lidars.…”

Section: Introductionmentioning

confidence: 99%

“…The key challenges in tactile sensing based control are the difficulty to accurately model physical contacts, partial observability of the environment from touch only, and noise in perception and control. Tactile feedback based manipulation controllers have been proposed, but they often use heuristic search patterns [5] or are tailored for a specific task [3]. In such case, the learning-based approaches are more promising to learn generic solutions for contract-rich manipulation control tasks.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Seq2Seq Imitation Learning for Tactile Feedback-based Manipulation

Yang¹,

Angleraud²,

Pieters³

et al. 2023

Preprint

View full text Add to dashboard Cite

Robot control for tactile feedback based manipulation can be difficult due to modeling of physical contacts, partial observability of the environment, and noise in perception and control. This work focuses on solving partial observability of contact-rich manipulation tasks as a Sequence-to-Sequence (Seq2Seq) Imitation Learning (IL) problem. The proposed Seq2Seq model first produces a robot-environment interaction sequence to estimate the partially observable environment state variables, and then, the observed interaction sequence is transformed to a control sequence for the task itself. The proposed Seq2Seq IL for tactile feedback based manipulation is experimentally validated on a door-open task in a simulated environment and a snap-on insertion task with a real robot. The model is able to learn both tasks from only 50 expert demonstrations while state-of-the-art reinforcement learning and imitation learning methods fail.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Seq2Seq Imitation Learning for Tactile Feedback-based Manipulation

Yang¹,

Angleraud²,

Pieters³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In 1948, Shannon quantitatively described the uncertainty of random event by the concept the entropy [3]. Information entropy is used to measure the uncertainty of random event and introduced to the maximum entropy reliability theory.…”

Section: The Maximum Entropy Reliability Theorymentioning

confidence: 99%

The Method Used for Process Precision Prediction of Micro-Assembly Based on Principle of Maximum Entropy

Zhang

Sun

et al. 2012

AMR

View full text Add to dashboard Cite

This paper briefly introduces the principle of sub-micro precision micro-assembly system based on the micro force detection. The linear mapping of micro force-micro displacement can be obtained, which is in agreement with the experimental data. And the probability distribution density function of the fitted mapping coefficients can be derived by the principle of maximum entropy. Furthermore, the uncertainly of k is accessed. Hence, the reliability of micro force-micro displacement mapping in the assembly process is demonstrated. The experiment result shows that the assembly precision can be accurately predicted by the fitted micro force-micro displacement mapping.

show abstract