Abstract:

"Intelligent robotic systems are increasingly crucial across diverse fields, including industrial operations, medical applications, nuclear facilities, and challenging environments. As advancements in health technology continue, the use of vital signals monitoring in complex and hazardous settings is expected to play a key role in supporting people’s daily activities, paving the way for safer and more efficient environments. The MARCHESE project (Machine Learning-based Human Recognition and Health Monitoring System) aims to extend the concept of contactless monitoring from industrial and emergency settings from the CERN accelerators complex with search and rescue robotics to healthcare and hospital environments. The system seeks to introduce a contactless approach for continuous, non-invasive health monitoring of patients by measuring physiological parameters (heart rate, respiratory rate, and body temperature) while also tracking non-physiological indicators (facial expressions, body tracking). Implementing this contactless monitoring system in healthcare settings is essential for enabling non-invasive hospital admissions, providing real-time alerts, tailoring patient-specific monitoring, and enhancing the responsiveness of medical staff. Additionally, remote monitoring minimizes direct physical interaction, significantly reducing the risk of infection transmission and ensuring the safety of both medical personnel and patients. Ongoing collaboration with the Hospital of Lille (CHU Lille) and the Campus Bio-Medico University of Rome (UCBM) has enabled the project to gain an in-depth understanding of clinical needs and conduct preliminary hospital-based trials. By incorporating continuous feedback from medical and hospital staff, the system is continuously updated to integrate the monitoring of non-physiological factors with physiological data, thereby offering a complete health assessment. The system currently utilizes a hardware design tailored to meet the specific requirements and constraints of hospital environments, combined with algorithms and machine learning models for accurately measuring and assessing physiological parameters alongside non-physiological indicators to provide a comprehensive evaluation of the patient’s condition. Preliminary tests conducted on healthy subjects in hospital settings have shown promising potential for improving patient outcomes and enhancing healthcare delivery efficiency. These tests have also identified departments that could benefit most from the technology, such as pediatrics, geriatrics, and palliative care."