CARMA

Collaborative Autonomous Robots for eMergency Assistance

CARMA aims to co-create, through a user-centred iterative methodology involving a complementary set of end-users and social sciences and human (SSH) experts, a groundbreaking, modular and intuitive platform offering a complementary set of semi-autonomous and autonomous Unmanned Ground Vehicles (UGVs) capable of working in symbiosis with humans to support and supplement first responders and assist citizen in a wide range of disaster situations, including those with very low visibility. The project will build on the most advanced research results, including those from the INTREPID project, in the field of disaster robotics making them autonomous thanks to novel 3D radar-based environment mapping and analysis combined with Artificial Intelligence (AI) for enhanced path and mission planning as well as victim and threat detection. Symbiotic operations and natural robot/human interaction will be made possible by exploiting Generative Adversarial Networks (GANs) for collaborative tasks, Natural Language Processing (NLP), and eXtended Reality (XR) technologies. The project will evaluate the effectiveness and acceptance of the proposed platform in the frame of four ambitious pilots in complementary operational environments. Social, ethical and legal constraints will be carefully considered during the project’s lifetime. In order to prepare the ground for broad adoption and successful exploitation of the project’s results, CARMA will implement an ambitious communication and dissemination plan, an engaging training curriculum, and produce a white book proposing recommendations for doctrine changes to involve the proposed solution in crisis management operations.

M4D will participate in the development and integration of smart cognition capabilities for the deployed UGVs. Specifically, M4D will leverage the power of the CARMA sensing module, that is, data captured from various sources, e.g., LiDAR, optical/thermal/depth sensors, IMUs, etc., to develop a deep learning fusion module for detecting and localizing paramount information for first responders and navigation purposes, such as obstacles, human in distress, hazard sources, etc. Furthermore, M4D, through leading the relevant task, will contribute in the definition and continuous update of the technological and standardization framework.