PACT: Proactive Asking for Continual Task Assistance in Human-Robot Collaboration

2026-05-22

AI Summary

This work addresses the challenge of providing efficient assistance in long-term human-robot collaboration, where robots initially lack knowledge of user habits and operate in partially observable environments. To overcome this limitation, the paper introduces PACT, a novel framework that incorporates an active querying mechanism into sustained collaborative settings. PACT dynamically decides whether to solicit clarification from the user by evaluating the sufficiency of the current contextual information to infer intent accurately. Implemented via reinforcement learning, the framework optimizes decision-making over multi-day embodied tasks and introduces a clarification utility metric to quantify the trade-off between assistance accuracy and query frequency. Experimental results demonstrate that PACT significantly outperforms passive inference baselines in both assistance accuracy and clarification utility, underscoring the critical role of active querying in long-term collaborative scenarios.

Abstract

Robotic assistants in long-term human-robot collaboration need to assist users under partial observations while leveraging cross-day interaction history. However, human traits and routines are often unknown at the beginning of collaboration, making passive infer-then-act assistance ineffective and inefficient. To address this challenge, we study a cross-day proactive asking setting for continual task assistance and propose PACT (Proactive Asking for Continual Task Assistance), an ask-or-act framework that determines whether clarification should be sought before taking action. PACT leverages current observations together with accumulated interaction history to evaluate contextual sufficiency, enabling the robot to provide more reliable assistance and progressively adapt to the user over time. We implement its primary learned instantiation using reinforcement learning and evaluate alternative instantiations under the same framework. To assess such behavior, we further introduce a clarification utility metric that quantifies the trade-off between assistance accuracy and the frequency of clarification requests. Experiments in multi-day embodied collaboration scenarios demonstrate that, compared with passive inference baselines, PACT consistently improves both assistance accuracy and clarification utility, highlighting the importance of proactive asking in continual human-robot collaboration.

Problem

• human-robot collaboration
• continual task assistance
• proactive asking
• partial observability
• interaction history

Innovation

• proactive asking
• continual task assistance
• human-robot collaboration
• clarification utility
• reinforcement learning