HumAMI kicks off in Eindhoven with the University of Sassari as a project partner

The HumAMI European project held its kick-off meeting on 21–22 September 2026 at Eindhoven University of Technology (TU/e) in the Netherlands. The project focuses on developing artificial intelligence for machines, robots, and systems that interact with their physical surroundings and with people.

The University of Sassari contributes to the project through its Department of Engineering and Department of Veterinary Medicine, with involvement across all Work Packages. The Department of Engineering leads WP3 on edge computing and embedded AI hardware and coordinates RTD11, a demonstrator focused on intelligent monitoring for precision livestock farming.

The two-day meeting brought the consortium partners together to discuss the project’s objectives, management structure, and planned activities. The first day featured introductions to the partners and Work Packages, alongside a session on ethics. The second day focused on the twelve demonstrators that will be used to test and validate the project’s technologies in practical settings.

HumAMI aims to strengthen Europe’s AI ecosystem by advancing Physical AI, which refers to artificial intelligence embedded in machines and systems that sense, interpret, and act on their physical environment. The project combines edge AI, which enables data processing directly on devices, with agentic AI, which supports reasoning, learning, planning, and decision-making. These technologies will be integrated with communication networks, cloud resources, and application services, with safety, reliability, explainability, and sustainability guiding their design. Applications will span mobility, energy, digital industry, healthcare, agrifood, and digital society.

Within this framework, RTD11, “Human-Inspired Monitoring System for Precision Farming”, coordinated by UNISS, will address key challenges in extensive small ruminant farming. These include limited opportunities to assist animals during lambing, predator attacks, and the effects of extreme weather. The aim is to identify abnormal behaviour and emerging risks early, helping farmers respond promptly and prevent avoidable losses while improving animal welfare and farm viability.

The monitoring system will combine wearable sensors and smart cameras to observe livestock in the field. Edge and fog computing will allow data to be processed on devices and nearby computing nodes, reducing reliance on cloud services. This approach is particularly relevant in rural areas with limited internet connectivity and will also help reduce energy use. AI algorithms, including computer vision methods inspired by human perception, will analyse sensor data to alert farmers and, where feasible, trigger automated responses.

The demonstrator will also incorporate digital twins, virtual models updated with sensor data to represent animal behaviour and health and support early intervention. Collaboration between the two Departments will bring together engineering expertise and the veterinary knowledge needed to interpret the data and assess the proposed solutions in practice.

The UNISS team includes Pietro Ruiu, the University’s main contact for the project, Mauro Fadda, Claudio Rubattu, and Andrea Lagorio from the Department of Engineering, and Fiammetta Berlinguer, Francesca Sotgiu, and Valeria Pasciu from the Department of Veterinary Medicine.

HumAMI is funded under the HORIZON-JU-Chips-2025-RIA-two-stage call, Grant Agreement No. 101252558.