Developing and testing of control software framework for autonomous ground vehicle

Automation in ground vehicles has been gaining momentum in recent years highlighted by the significant number of public demonstrations in the last two decades. This momentum has created an urgency within research organizations, vehicle manufacturers and academia to solve existing problems with autonomous vehicle technology to make it usable in the real world. As autonomous ground vehicles operate in close proximity to one another, the margin of error for navigation is smaller than in other domains such as aerospace and marine application. The real-world driving scenarios for the autonomous ground vehicle can sometimes be predictable and unpredictable at other times, demanding different behaviours from the autonomous vehicle for successful navigation. To satisfy such as requirement, the autonomous vehicle should exhibit the capability to adapt to through deliberative planning in predictable environments and reactive planning in unpredictable environments. In this paper, we describe a hybrid control software framework designed to incorporate behaviour planning algorithms that are capable of both deliberative and reactive planning. The paper describes the development of this novel adaptive autonomous control software framework and validates it through both virtual testing and real world testing environments