SPOC: Safety-Aware Planning Under Partial Observability And Physical Constraints
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912926139416576 |
|---|---|
| author | Kim, Hyungmin Jeon, Hobeom Kim, Dohyung Jang, Minsu Kim, Jeahong |
| author_facet | Kim, Hyungmin Jeon, Hobeom Kim, Dohyung Jang, Minsu Kim, Jeahong |
| contents | Embodied Task Planning with large language models faces safety challenges in real-world environments, where partial observability and physical constraints must be respected. Existing benchmarks often overlook these critical factors, limiting their ability to evaluate both feasibility and safety. We introduce SPOC, a benchmark for safety-aware embodied task planning, which integrates strict partial observability, physical constraints, step-by-step planning, and goal-condition-based evaluation. Covering diverse household hazards such as fire, fluid, injury, object damage, and pollution, SPOC enables rigorous assessment through both state and constraint-based online metrics. Experiments with state-of-the-art LLMs reveal that current models struggle to ensure safety-aware planning, particularly under implicit constraints. Code and dataset are available at https://github.com/khm159/SPOC |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_21595 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | SPOC: Safety-Aware Planning Under Partial Observability And Physical Constraints Kim, Hyungmin Jeon, Hobeom Kim, Dohyung Jang, Minsu Kim, Jeahong Robotics Embodied Task Planning with large language models faces safety challenges in real-world environments, where partial observability and physical constraints must be respected. Existing benchmarks often overlook these critical factors, limiting their ability to evaluate both feasibility and safety. We introduce SPOC, a benchmark for safety-aware embodied task planning, which integrates strict partial observability, physical constraints, step-by-step planning, and goal-condition-based evaluation. Covering diverse household hazards such as fire, fluid, injury, object damage, and pollution, SPOC enables rigorous assessment through both state and constraint-based online metrics. Experiments with state-of-the-art LLMs reveal that current models struggle to ensure safety-aware planning, particularly under implicit constraints. Code and dataset are available at https://github.com/khm159/SPOC |
| title | SPOC: Safety-Aware Planning Under Partial Observability And Physical Constraints |
| topic | Robotics |
| url | https://arxiv.org/abs/2602.21595 |