Research paper
Safety benchmarks for planning in unseen space
The paper compares safe policies for robots that sense an unknown environment incrementally. It develops an optimality benchmark and a tractable policy class that can exploit learned priors, evaluated numerically across different dynamics.
Reviewed 2026-11-05 · Source and bibliography checked
- Original title
- Safe Motion Planning in Unknown Environments: Optimality Benchmarks and Tractable Policies
- Publisher
- Robotics: Science and Systems Foundation
- Authors
- Lucas Janson, Tommy Hu, Marco Pavone
- DOI
- 10.15607/rss.2018.xiv.061
- Proceedings publication date
- 2018-06-26
- Source language
- en
- Edition
- Robotics: Science and Systems XIV
- Access
- Public bibliography and abstract
- Publication status
- Published
Source overview
The paper compares safe policies for robots that sense an unknown environment incrementally. It develops an optimality benchmark and a tractable policy class that can exploit learned priors, evaluated numerically across different dynamics.
Use and review
The reviewed evidence is numerical or simulation based. Physical robot behavior, perception failures and implementation timing require separate validation.
Sources & editions
Safe Motion Planning in Unknown Environments: Optimality Benchmarks and Tractable Policies ↗
Robotics: Science and Systems Foundation · Robotics: Science and Systems XIV
Owning RSS2018 title/authors/DOI/abstract/BibTeX checked. Numerical examples only in reviewed abstract.