Jump to content

Verifiable and safe AI for Autonomous Systems

Verifiable and safe AI for Autonomous Systems

The rapidly growing application of machine learning techniques in cyber-physical systems leads to better solutions and products in terms of adaptability, performance, functionality and usability.

Verifiable and safe AI for Autonomous Systems

Verifiable and safe AI for Autonomous Systems

The rapidly growing application of machine learning techniques in cyber-physical systems leads to better solutions and products in terms of adaptability, performance, functionality and usability.

However, cyber-physical systems are often safety-critical, and the resulting need for verification against potentially fatal accidents is self-evident and of crucial importance.

This project aims to develop easy-to-use verification methods and tools for safety-critical systems. 

Relevant domains include all types of autonomous systems where machine learning is applied, including water management.

The researchers use UPPAAL Stratego, a further development of the tool UPPAAL used for improving, validating and constructing systems. The tool was developed at the Department of Computer Science in collaboration with Uppsala University – and is used in several applications – from controlling traffic and floor-heating systems to developing better decision-making tools for COVID-19 initiatives.

Contact

Professor Kim Guldstrand Larsen

Professor with Specific Responsibilities Thomas Dyhre Nielsen

Facts

About Verifiable and safe AI for Autonomous Systems
The project is funded by Innovation Fund Denmark
Partners
  • ITU
  • Grundfos
  • HOFOR
  • Seluxit
  • Aarhus Vand
Project period
2021-2024