HumAMI aims to strengthen the European AI ecosystem by establishing a robust Physical AI framework. Physical AI refers to AI

solutions that operate within and interact with real-world physical systems such as machines, robots, and human-centric cyber-physical systems (CPSs). A Physical AI embodiment is such a machine, robot, or CPS with advanced machine intelligence (AMI). A

Physical AI embodiment that interacts with humans in the loop inherently requires human-inspired characteristics – adaptability,

explainability, trustworthiness, ethics, and sustainability – to be present in its entire design trajectory and lifecycle to ensure safe,

secure, and robust operation. To achieve this, the HumAMI framework brings together (i) edge AI, and (ii) agentic AI enabling AMI

such as reasoning, learning, action, and decision-making. Transitioning towards Physical AI also requires supporting (i) cloud

resources, (ii) connectivity, and (iii) applications & services. The HumAMI framework encompasses these by introducing a 6-layer

conceptual framework composed of – Physical infrastructure, Connectivity, Embedded AI, Agentic AI, Cloud platform, and

Applications & Services, to realize Physical AI embodiment in industrial machines, robot,s and human-centric CPSs.

HumAMI framework is realised by (i) developing fundamental innovative disruptive technologies (IDTs) (open science objective

openAMI), (ii) developing integrated technology solutions (ITSs) bringing multiple IDTs together (cooperation objective coopAMI),

and (iii) demonstrating and validating IDTs and ITSs in real-world technology demonstrators (RTDs), fostering industrial adoption

(adoptAMI objective) in mobility, energy, digital industry, health, agrifood, and digital society. The HumAMI initiative will significantly

enhance European industrial competitiveness, promote the development of ethical AI, and secure Europe's leadership in the global

Physical AI landscape, balancing technological advancement with core societal values.