European partners launch an equity-free accelerator for physical AI startups

UnternehmerTUM Accelerator, Microsoft, Infineon Technologies and NVIDIA announced the Physical AI Growth Lab on October 22, opening applications for an equity-free programme aimed at robotics and industrial AI startups. The advertised first programme runs from January to April 2027. Its launch is a new institutional offer to founders, not evidence that participating companies have already completed it or achieved commercial scale.

The original Infineon announcement dates the launch in Munich and identifies UnternehmerTUM Accelerator as the programme lead. Independent Dutch industrial publication Aandrijven & besturen corroborates the partnership, application deadline and planned format.

A programme for teams with working prototypes

The stated target is 10–15 teams based in Europe, the Middle East or Africa, or building products for those markets. Applicants should have a working prototype or minimum viable product, at least one technical co-founder, a team of no more than twenty people and a company founded within the previous three years. Applications are due by November 13, 2026.

The partners describe a hybrid format combining online activity with in-person sessions in Munich. The offer includes development resources, mentoring and access to industry customers and investors without taking equity in participating companies. That programme term does not determine the terms of any separate financing a startup might later seek.

The resources span several parts of a robotics business: cloud infrastructure, AI software and simulation, semiconductor hardware, and business development. The independent report describes the same combination. The relevant news is that these organisations are packaging support around physical systems rather than offering only a software demonstration.

What the launch does not establish

Robotics companies must connect models to sensors, controllers and hardware that work under actual customer conditions. Access to tools and contacts may help with that process, but it does not guarantee that a prototype becomes a dependable product or that an investor will fund it.

The checked announcements do not name a completed cohort or provide measured results from this future programme. Selection decisions, customer contracts, product reliability and subsequent financing remain outcomes to observe. The partners’ intention to accelerate commercial development should therefore stay distinct from evidence of commercial success.