SmartWalkCoach: An AI Companion for End-to-End Walking Guidance, Motivation, and Reflection

2026-05-14

AI Summary

This study addresses a critical gap in existing walking assistance tools, which often fail to seamlessly integrate navigation support with motivational engagement while maintaining low cognitive load throughout the walking experience. To bridge this gap, the authors propose the first tool-augmented multi-agent architecture that spans the entire walking cycle. The system integrates map APIs, context-aware dialogue generation, and behavior intervention strategies through three lightweight agents that collaboratively enable interest-driven route planning, timely affective encouragement during walking, and post-walk reflection and planning. A field crossover study (N=12) demonstrates that the system significantly enhances users’ positive affect and overall experience. The work further distills key design principles for mobile health companions, emphasizing supportive relational expressions and strategic timing of interventions to avoid high cognitive load.

Abstract

We present SmartWalkCoach, a mobile AI companion that supports the full walking journey: from pre-walk planning to in-walk guidance through to post-walk reflection. Addressing a gap between map navigation and motivational coaching, SmartWalkCoach orchestrates three lightweight agents: (1) GeographyAgent for conversational route curation from nearby points of interest and user preferences while delegating pathfinding to map APIs; (2) AccompanyAgent for context-aware, just-in-time prompts that blend informational cues with relational encouragement; and (3) SummaryAgent for concise reflection and next-step planning. This end-to-end, tool-using design aims to lower cognitive load in planning and sustain engagement and motivation during walking through delivering dynamic, cadence-aware interventions. We conducted an in-the-wild, two-period AB/BA crossover study (N=12), where each participant completed two comparable walks with counterbalanced conditions: Information-only versus Information+Motivation. Linear mixed models show that adding motivational, companion-like dialogue significantly improved outcomes: participants reported higher positive feelings and better user experience, with no evidence of carryover. Thematic analysis surfaced two design imperatives for mobile companions: supportive, relational expression and context-aware timing (e.g., avoiding high-load moments, intervening at fatigue/milestones). Our contributions are: (i) an end-to-end, tool-using agent architecture for everyday walking that reduces cognitive load during planning and accompaniment; (ii) a controlled field evaluation linking context-aware motivation to affect and UX gains; and (iii) actionable design guidance on expression, timing, and frequency for mHealth companions.We outline limitations and paths toward multimodal, voice-first companions, with adaptive personalization mechanisms.

Problem

• walking guidance
• motivational coaching
• mobile AI companion
• cognitive load
• context-aware intervention

Innovation

• AI companion
• context-aware motivation
• multi-agent architecture
• mobile health
• just-in-time intervention