LATIDIA · Robótica

Colaboración heterogénea de robots en entornos no estructurados con inteligencia generativa fundamentada

arXiv:2510.26915v2 Tipo de anuncio: reemplazar Resumen: Si bien los equipos heterogéneos generalmente se han diseñado para misiones bien especificadas con semántica conocida, inteligencia generativa, es decir, modelos de lenguaje grande (LLM) y

La noticia

arXiv:2510.26915v2 Announce Type: replace Abstract: While heterogeneous teams have typically been designed for well-specified missions with known semantics, generative intelligence, i.e., large language models (LLMs) and vision language models (VLMs), opens the possibility of teams that infer mission-relevant semantics and subtasks given high-level natural language specifications and environmental context. However, current LLM- and VLM-enabled teaming methods typically assume well-structured and known environments, limiting performance in complex real-world settings. We address these limitations via SPINE-HT, a framework that grounds the reasoning abilities of LLMs in the evolving context of a heterogeneous robot team through a three-stage process. Given mission specifications and team capabilities in natural language, an LLM infers necessary subtasks.