Artificial Wingman and Co-pilot

Important results from the project

This project explored the realms of machine learning applications within the domain of air combat, with a particular focus on future applications. Initially conceived as an exploration into fully autonomous systems, the focus shifted towards developing decision support mechanisms. This shift was motivated by the versatility of decision support, which can cater to both pilots and autonomous aircraft. The research contributed to academic advancement in the field and fostered industrial innovation.

Expected long term effects

Throughout the project, significant progress was made on the research front, including publications in prestigious venues such as the IEEE Control Systems Letters journal, ICRA and ICUAS conferences. An open-source Air Combat Beyond Visual Range was also developed to foster collaboration and facilitate experimentation within the research community. Fighter pilots at Saab have given positive feedback on our ideas and results. The proof of concept is being considered to be implemented at Saab.

Approach and implementation

The industrial Ph.D. student had a master´s degree in aeronautical engineering, but had to study machine learning from scratch for the research assignment. This has gone well, and he has produced several strong publications in the field. The Ph.D. student has developed a decision support tool to help operators in Beyond Visual Range (BVR) air combat scenarios assess the risks of different options and make decisions based on those. The results have been presented at scientific conferences.