UrbanHuRo: A Two-Layer Human-Robot Collaboration Framework for the Joint Optimization of Heterogeneous Urban Services

A research team from undisclosed institutions has introduced UrbanHuRo, a novel framework for coordinating human couriers and sensing robots to simultaneously optimize multiple urban services. Accepted as a regular paper at the prestigious International Conference on Robotics and Automation (ICRA'26), the system addresses a critical gap in smart city research by moving beyond isolated service optimization to consider reciprocal interactions between heterogeneous services. The framework demonstrates how human couriers can collect traffic and air quality data during deliveries while robots assist with on-demand delivery during peak hours, creating a symbiotic relationship that enhances both sensing coverage and delivery efficiency.

The technical implementation features two key innovations: a scalable distributed MapReduce-based K-submodular maximization module for efficient order dispatch, and a deep submodular reward reinforcement learning algorithm for dynamic sensing route planning. Experimental evaluations using real-world datasets from a food delivery platform showed remarkable results, with UrbanHuRo improving sensing coverage by 29.7% and increasing courier income by 39.2% on average across most settings. The system also significantly reduced the number of overdue orders, demonstrating practical benefits for both service providers and workers. This research represents a significant step toward more integrated urban service ecosystems where human and robotic agents collaborate rather than operate in isolation.