Safety-aware planning & control
Linking reasoning and task plans to whole-body constraints, balance, and contact-aware action.
Full-day workshop
From Capability
to Assurance:
Trustworthy Humanoid Systems
in Human Environments
Human environments. Trustworthy systems.
01 The motivation
Assurance across the humanoid system.
Humanoid safety requires more than safe AI models or safe robot control considered in isolation. Learning-enabled humanoids connect perception, reasoning, planning, whole-body action, physical interaction, runtime supervision, and recovery.
Errors or uncertainty at one layer can propagate into unsafe contact, balance loss, object dropping, or inappropriate whole-body actions around people. This workshop examines how safeguards work across these layers and what evidence can demonstrate their effectiveness under uncertainty and failure.
Central technical questions
How can humanoids perceive, reason, plan, act, and recover safely under uncertainty and failure, and how do we evaluate those safeguards?
What evidence is sufficient to trust a learning-enabled humanoid operating physically around people?
02 Topics of interest
The foundations of assurance.
Verification, validation & assurance
Testing assumptions across learned components, controllers, and runtime safeguards; building system-level evidence.
Human–humanoid interaction
Safe contact, object handling, and predictable whole-body behavior around people.
Runtime supervision & recovery
Detecting uncertainty and failures, supervising learned actions, and recovering from unsafe states.
Benchmarks & evaluation
Evaluating error propagation, physical consequences, and safeguard effectiveness in simulation and real-world tests.
Responsible deployment
Defining operational boundaries and connecting monitoring and incident evidence to lifecycle assurance.
03 Contribute to the workshop
Call for Papers & Demos
We invite papers and demos addressing safety and assurance for humanoid systems operating in human environments.
We welcome two types of paper submissions:
Short / Position Papers — up to 4 pages
Early-stage research, position papers, preliminary results, failure analyses, benchmark proposals, safety architectures, and emerging research directions.
Research Papers — 4–8 pages recommended
Substantially developed research contributions, including methods, system studies, experiments, benchmarks, and safety evaluation studies.
Paper submission information
- Format
- Paper submissions should use the IEEE standard double-column conference format. One additional page may be used exclusively for references. Appendices are not considered as part of the review submission.
- Review process
- Papers will undergo single-blind review. Authors do not need to anonymize their submissions. Review notifications will be sent out on a rolling basis from November 2 to November 20. Papers will be reviewed as submissions are received, and we will make every effort to notify authors of the review results as early as possible.
- Presentation
- Accepted papers will be presented as posters. Selected papers will additionally be invited for short spotlight presentations.
- Publication policy
- No formal proceedings (we are a non-archival workshop). Authors are welcome to post their papers on arXiv.
Call for Demos
Demo Submit!
Demo submissions are now open! The deadline is November 15, 2026 (23:59 AoE). Paper and demo submissions use the same OpenReview portal.
We welcome demos using simulation, portable equipment, or physical robots.
- File format
- Any file format is accepted.
- Title / heading
- The title or heading of your submitted document must clearly include the word Demo.
- Demo materials
- Include a demo video or animation and a brief description of the demo's functionality. Clearly state whether the demo uses simulation, portable equipment, or a physical robot.
- Equipment
- Applicants are generally expected to bring their own demo equipment. Please specify any additional requirements in your submission.
- Presentation arrangements
- Details of the demo presentation format, schedule, and on-site setup arrangements will be announced.
- Special arrangements & questions
- If you have any special requests, questions, or need additional arrangements, please feel free to contact us directly via email.
05 Organizing committee
Organizers
Chuchu Chen
Assistant Professor
George Washington University
Georgios Fainekos
Senior Principal Scientist
Toyota Motor North America
Abhijeet Kulkarni
Ph.D. Candidate
University of Delaware
Hideki Okamoto
Senior Principal Engineer
Toyota Motor North America
Zijun Sha
Research Engineer
Toyota Motor Corporation
06 Get in touch
Contact
For questions about contributions, participation, special demo requests, or additional arrangements, please feel free to contact us directly via email.