Can AI read sensor signals?

Can AI read sensor signals? D-MTEC spin-off wins major EU funding

A doctor deciding whether a patient needs urgent care. A factory team catching a machine fault before the line stops. A grid operator judging which alarm is real. These decisions all rest on the same thing: time series data, the continuous streams of numbers produced by sensors and monitoring systems. Most AI language models are built to read text. Aionic Labs, a new company founded on research from the Agentic Systems Lab at D-MTEC, is teaching AI to read signals instead.

Aionic Lab

From lab to launch

The start-up is led by Dr. Robert Jakob, Director of the Agentic Systems Lab at ETH Zurich, together with co-founder Thomas Kaar. Their technology, called Time-Series Language Models (TSLMs), analyses sensor and measurement data – industrial machinery, patient vitals, power grid loads, banking and logistics data – and explains what it means in plain language.

The approach grew out of an open research collaboration between ETH Zurich, Stanford University, and Google, published as OpenTSLM and presented at the International Conference on Machine Learning (ICML), one of the field's top venues.

Backed by €26.5 million

Aionic Labs has now been selected by SPRIN-D, Germany's Federal Agency for Disruptive Innovation, for its "Next Frontier AI" programme – one of only ten teams chosen. The 24-month programme is designed to turn promising AI research into deployable technology. Aionic Labs will receive around €3 million in the first stage, rising to as much as €26.5 million if it advances through subsequent stages.

The programme's broader goal is to help build a European frontier AI lab and reduce the continent's dependence on foreign AI models.

What's next

Based in Zurich, Aionic Labs plans to launch pilot projects across manufacturing, healthcare, energy, logistics, telecoms, insurance and banking over the coming months, and to grow an open research community around OpenTSLM.

For D-MTEC, the launch is a concrete example of research moving from publication to practice – and of a department lab shaping how European industry works with AI.

More information and open positions: external page aioniclabs.ai

 

JavaScript has been disabled in your browser