AI-based Tactical Mission Planner

Artificial intelligence will make it possible to reduce the workload on pilots and considerably increase mission efficiency by offering decision support on board aircraft through intelligent assistants.

Carrying out a mission to find the best trajectory in enemy territory can be an extremely complex task for a helicopter or a fighter pilot. In addition to control your aircraft, you must take into account multiple constraints such as terrain, surface radars or ground/air threats.

Thanks to artificial intelligence and our operational expertise, Thales is developing an assistance function that the pilot will be able to use directly in flight to know which are the optimized flight plans taking into account all operational constraints and meeting the objectives of mission in terms of efficiency or risk-taking

Thales has designed the TMP solution to make the collaboration process between humans and intelligent functions more efficient. It is designed to assist humans based on high-level intentions and priorities alone, rather than multiple and complex cockpit interface setting.

In the warfare context, mission intention can be associated to mission objectives or mission priorities (high-risk mission, high effectivity mission) imposed by the air mission commander. Therefore, humans and intelligent system can collaborate through the appropriate level of abstraction naturally used by humans depending on the context reducing cognitive workload in mission management.

The TMP is based on AI technologies specifically adapted for critical systems. It is a smart mix of advanced algorithms and in-house GFT (Genetic Fuzzy Tree) that provide fully interpretable and transparent solutions because it is designed using expert knowledge as well as data. That means that humans can easily understand what the TMP is providing as an output or recommendation with real advantage compared to some other AI technologies that are not or approximately explainable such as neural network.

Based on operational expertise and optimized through machine learning, TMP is providing the best compromise between explainability and scalability. It builds trust with users wanting to solve very complex problems in mission critical decision-making applications.

Moreover, TMP solution is formally verifiable once trained against upper-level properties or requirements guaranteeing that TMP does not violate any safety or mission critical requirements such as aircraft performance limitations or rules of engagement.