English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Policy learning for delicate industrial insertion tasks (e.g., PC boardassembly) is challenging. This paper considers two major problems: how to learna diversified policy (instead of just one average policy) that can efficientlyhandle different workpieces with minimum amount of training data, and how tohandle defects of workpieces during insertion. To address the problems, wepropose tolerance-guided policy learning. To encourage transferability of thelearned policy to different workpieces, we add a task embedding to the policy'sinput space using the insertion tolerance. Then we train the policy usinggenerative adversarial imitation learning with reward shaping (RS-GAIL) on avariety of representative situations. To encourage adaptability of the learnedpolicy to handle defects, we build a probabilistic inference model that canoutput the best inserting pose based on failed insertions using the tolerancemodel. The best inserting pose is then used as a reference to the learnedpolicy. This proposed method is validated on a sequence of IC socket insertiontasks in simulation. The results show that 1) RS-GAIL can efficiently learnoptimal policies under sparse rewards; 2) the tolerance embedding can enhancethe transferability of the learned policy; 3) the probabilistic inference makesthe policy robust to defects on the workpieces.

Tolerance-Guided Policy Learning for Adaptable and Transferrable Delicate Industrial Insertion