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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Grasping small objects surrounded by unstable or non-rigid material plays acrucial role in applications such as surgery, harvesting, construction,disaster recovery, and assisted feeding. This task is especially difficult whenfine manipulation is required in the presence of sensor noise and perceptionerrors; errors inevitably trigger dynamic motion, which is challenging to modelprecisely. Circumventing the difficulty to build accurate models for contactsand dynamics, data-driven methods like reinforcement learning (RL) can optimizetask performance via trial and error, reducing the need for accurate models ofcontacts and dynamics. Applying RL methods to real robots, however, has beenhindered by factors such as prohibitively high sample complexity or the hightraining infrastructure cost for providing resets on hardware. This workpresents CherryBot, an RL system that uses chopsticks for fine manipulationthat surpasses human reactiveness for some dynamic grasping tasks. Byintegrating imprecise simulators, suboptimal demonstrations and external stateestimation, we study how to make a real-world robot learning system sampleefficient and general while reducing the human effort required for supervision.Our system shows continual improvement through 30 minutes of real-worldinteraction: through reactive retry, it achieves an almost 100% success rate onthe demanding task of using chopsticks to grasp small objects swinging in theair. We demonstrate the reactiveness, robustness and generalizability ofCherryBot to varying object shapes and dynamics (e.g., external disturbanceslike wind and human perturbations). Videos are available athttps://goodcherrybot.github.io/.

Cherry-Picking with Reinforcement Learning : Robust Dynamic Grasping in Unstable Conditions