HygieneRoboBench: Benchmarking Hygiene-Aware Planning for Household Robots
Yurun Chen, Josh Qixuan Sun, Jason Qin, Chengtai Li, Tianyi Wang, Mark Crowley, Wentao Zhu
- Published
- Oct 6, 2026 — 16:33 UTC
Problem
The paper identifies a significant gap in the assessment of hygiene risks and safe planning for household robots. Existing frameworks do not adequately address the complexities involved in hygiene-aware task execution, particularly in domestic environments. This work is presented as a preprint and has not undergone peer review.
Method
The authors propose HygieneRoboBench, a comprehensive benchmark consisting of 624 instances across 134 task families. The benchmark evaluates tasks that involve capturing contamination through two grippers and shared objects, while also considering treatment costs and user priorities. The evaluation method employs controlled history, profile, and event comparisons, allowing for independent plan evaluation. The proposed planning method, Hygiene-NSP, integrates large language model (LLM)-based grounding, contact-history reconstruction, and Constraint Programming (CP) with SAT solving for effective planning.
Results
The proposed Hygiene-NSP method achieves a Safe Resolution Rate of 94.4% and an Optimal Safe Resolution Rate of 90.4%, both compared against baseline planners. These metrics indicate a significant improvement in the ability of household robots to plan tasks while maintaining hygiene standards.
Limitations
The authors do not report any limitations in their study. However, the absence of reported limitations may suggest a need for further exploration of the benchmark's applicability across diverse household scenarios and the robustness of the proposed method under varying conditions.
Why it matters
This work has important implications for the development of household robots, particularly in enhancing their ability to operate safely in environments where hygiene is a concern. By establishing a benchmark for hygiene-aware planning, it paves the way for future research to build upon these findings, potentially leading to more sophisticated and reliable robotic systems that can effectively manage hygiene risks.
By Turing Wire Research Desk · Oct 6, 2026 · How we work →
Summarised from the paper by the Turing Wire Research Desk. The full paper has the complete methods and results.
Source: arXiv cs.AI
