A Platform for Designing Solutions for Automotive Autonomous Driving based on Evolutionary Robotics approach

We consider the autonomous driving of real vehicles as a challenging framework where Robotics-oriented approach might be effective in solve complex tasks. Here we discuss some of the autonomous navigation features in order to set up an environment to develop neural control systems for automotive autonomous driving. In this work we introduce the Evolutionary Robotics as a framework for automatically evolve and select neuro-controllers to test in a conceptual experiment reproducing the road scenario features. Preliminary experiments of Cruise Control and Adaptive Cruise Control were conducted in simulation, comparing the results of both our simulator and a commercial dedicated simulator providing a complete car library. The experiments help in evaluating the feasibility of novel approaches in the development of controllers for automotive vehicles.