Robots improve searching for people using habits and trust in real settings
CSIR: Contextually and Socially Informed Robots for Efficient Person Goal Navigation
Robotics
Summary
Finding people with robots can be tricky when the robot doesn’t have exact details about who to find. The authors developed a way for robots to search for people by combining simple distance measures with information about what people like to do and what they want, while considering how much the robot trusts the information given. They tested their method in computer simulations and on a real robot, showing it works better than traditional distance-based searching. Their approach could help robots interact naturally and efficiently with people in everyday environments.
What this means in practice
- •For service robot developers: Design robots that use habits and trust to find people efficiently in indoor environments with uncertain information.
- •For warehouse automation teams: Implement improved person-finding in indoor delivery robots to handle vague or partial information about recipients.
Authors
Tyler Chung, Nahl Farhan, Hao Zhang, Mingfeng Yuan, Jinjun Shan, Amy Wu, Yue Hu
Abstract
We address the Person Goal Navigation (PersonNav) problem, enabling robots to search for people under realistic constraints on information accessibility and language uncertainty. This tackles the current solutions revolving around rigid person finding systems requiring exact knowledge of the individual for item-delivery in indoor settings. While the focus in literature is on learned methods using unavailable public or sparse data due to human privacy. We propose a planning framework that combines distance and semantic information about the recipient (habits and intent) weighted by the trust of the user-provided information. A synthetic benchmark of scenarios is also developed including actors, items, and requests to evaluate performance before real-world deployment aiming to simulate natural language human-robot interactions. Results show our informed search outperforms classical distance-based graph baselines, while semantics alone lead to ungrounded, sporadic search. Our method achieves strong gains over baselines, with an LLM-based variant performing comparably, and its explicit belief representation naturally supports future Bayesian filtering. Hardware tests demonstrate our method supports real-world embodiment able to leverage between semantic and distance information in a real setting. This work moves toward more intelligent mobile service agents capable of human-like, informed search in realistic environments to be leveraged in day-to-day use. https://anonymous.4open.science/r/personnavsite-4ED2/index.html.