Using Artificial Intelligence to help decide the best combat tactics

In December 2017 artificial intelligence (AI) computer program AlphaZero beat the best chess, go and shogi (Japanese chess) games programs after just a couple hours of training itself. This inspired MBDA’s researchers to believe that “machine learning” algorithms and AI techniques could help improve the performance of weapon systems.

For example, with air defence, describing the behaviour of missiles (both offensively and defensively) to a “learning machine” could enable it to run a huge number of simulation scenarios representing realistic combat situations. Just like AlphaZero, this machine could very quickly become an expert in defence tactics and suggest the solution to counter an attack.

The objective of our research is not to replace human decision-making in a military confrontation; we will always need this. However, just like AlphaZero’s game playing, we would expect to discover defence tactics that are efficient against unforeseen strike scenarios, and which could give a human operator a decisive aid when confronted by complex attacks.

This type of research has led MBDA to be among nine companies as a sponsor in the ‘Challenge IA’ scheme. Part of the plan announced by the French President in March 2018 to boost artificial intelligence for the benefit of French economy, this scheme aims to promote open innovation initiatives between companies or public bodies and start-ups or small and medium enterprises (SMEs). The goal is the emergence of radically new solutions to the digital challenges faced.

To find out more visit:

> Press contact