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

A critical need in assistive robotics, such as assistive wheelchairs fornavigation, is a need to learn task intent and safety guarantees through userinteractions in order to ensure safe task performance. For tasks where theobjectives from the user are not easily defined, learning from userdemonstrations has been a key step in enabling learning. However, most robotlearning from demonstration (LfD) methods primarily rely on optimaldemonstration in order to successfully learn a control policy, which can bechallenging to acquire from novice users. Recent work does use suboptimal andfailed demonstrations to learn about task intent; few focus on learning safetyguarantees to prevent repeat failures experienced, essential for assistiverobots. Furthermore, interactive human-robot learning aims to minimize effortfrom the human user to facilitate deployment in the real-world. As such,requiring users to label the unsafe states or keyframes from the demonstrationsshould not be a necessary requirement for learning. Here, we propose analgorithm to learn a safety value function from a set of suboptimal and faileddemonstrations that is used to generate a real-time safety control filter.Importantly, we develop a credit assignment method that extracts the failurestates from the failed demonstrations without requiring human labelling orprespecified knowledge of unsafe regions. Furthermore, we extend ourformulation to allow for user-specific safety functions, by incorporatinguser-defined safety rankings from which we can generate safety level setsaccording to the users' preferences. By using both suboptimal and faileddemonstrations and the developed credit assignment formulation, we enablelearning a safety value function with minimal effort needed from the user,making it more feasible for widespread use in human-robot interactive learningtasks.

Credit Assignment Safety Learning from Human Demonstrations