SaltHawk: Self-learning Swarm Intelligence for Winter Road Service
Erich H. Franke, Roswitha M. Franke, Patrick Böer, Gerald Meyer, Xavier Groelly
Winter road service is an important issue in the field of traffic safety. It is also a cost-intensive endeavour, however, meaning that any optimisation is beneficial for both the environment and public expenditure. Currently, drivers select their dispenser’s settings on their own depending on their individual experience. Road conditions, temperatures, and so on can nonetheless vary locally and be subject to sudden changes. Since no global and centralised “boss” can be appointed to control everything, we decided to find a different approach. This led us to investigate how ants would behave if they were given this task, and thus was the SaltHawk swarm born. The idea behind SaltHawk involves forging a group of individual dispensers into a swarm of computers with local intelligence using an artificial intelligence approach and communicating over a secure and reliable network. The swarm learns from experienced drivers, weather, and road topology; each on-board computer knows its EGNOS/GPS location and decides on its own if and when to share information with the others.