Project Robin

Abstract

ROBIN (Robotic Assistance for Independent Nursing in High-Cost Diseases) is an industry-driven R&D project aimed at transforming
home-based care for chronic and high-cost diseases through an integrated, edge-centric robotic platform.

At the core of the solution is a social robotic system designed to interact naturally with patients, providing continuous support for
monitoring, daily care, and engagement. The robot integrates autonomous mobility, multimodal interaction (voice, vision, contextual awareness), and assistive functionalities, enabling safe operation in real home environments. It can detect anomalies, support treatment adherence, and facilitate communication with caregivers. Through embedded computer vision and environmental sensing, the robot builds a contextual understanding of its surroundings, enabling adaptive and personalized assistance.  The system relies on a decoupled architecture separating the robotic unit from a dedicated edge-based health controller, where most AI processing, data fusion, and decision-making are executed locally. This approach enables continuous, privacy-preserving monitoring while reducing latency, improving scalability, and allowing computational resources to evolve independently from the robotic platform.

This architecture is supported by advanced communication technologies, leveraging hybrid connectivity combining 5G, 5GAdvanced, emerging 6G networks, LPWAN, and non-terrestrial connectivity, ensuring reliable operation across urban, rural, and low-connectivity scenarios.

ROBIN introduces a distributed sensing and perception framework integrating computer vision, environmental sensing, and realtime localization. Localization is enhanced through 5G/6G Joint Communication and Sensing (JCAS), enabling precise indoor positioning and spatial awareness.

The platform integrates on-device AI containers and NPU-based acceleration to support real-time analytics and predictive models with minimal latency. A hybrid edge architecture fuses high-throughput sensing with low-bandwidth telemetry, enabling continuous operation and supporting digital twin representations for real-time robotic navigation and monitoring.

ROBIN will be validated through real-world clinical pilots, demonstrating impact in reducing hospitalizations, improving early detection, and enabling scalable care delivery.