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 search reinforcement learning allows robots to acquire skills bythemselves. However, the learning procedure is inherently unsafe as the robothas no a-priori way to predict the consequences of the exploratory actions ittakes. Therefore, exploration can lead to collisions with the potential to harmthe robot and/or the environment. In this work we address the safety aspect byconstraining the exploration to happen in safe-to-explore state spaces. Theseare formed by decomposing target skills (e.g., grasping) into higher rankedsub-tasks (e.g., collision avoidance, joint limit avoidance) and lower rankedmovement tasks (e.g., reaching). Sub-tasks are defined as concurrentcontrollers (policies) in different operational spaces together with associatedJacobians representing their joint-space mapping. Safety is ensured by onlylearning policies corresponding to lower ranked sub-tasks in the redundant nullspace of higher ranked ones. As a side benefit, learning in sub-manifolds ofthe state-space also facilitates sample efficiency. Reaching skills performedin simulation and grasping skills performed on a real robot validate theusefulness of the proposed approach.

Safe-To-Explore State Spaces: Ensuring Safe Exploration in Policy Search with Hierarchical Task Optimization