Hardware-in-the-Loop Platform for Testing Autonomous Vehicle Control Algorithms

Abstract

Autonomous vehicles are a demanding topic which has attracted a lot of attention lately. This kind of vehicle should be capable of sensing its environment and react on the presence of obstacles and other traffic participants. Algorithms for obstacle avoidance and overall control of the autonomous driving in relation to robustness regarding environmental conditions are required for driving safely on roads. For the development of these algorithms, a safe environment is needed. In order to speed up the development of the algorithms, this paper proposes a Hardware in the loop (HIL) based test environment and analysis of it for autonomous vehicle development. Proposed HIL is combining a virtual driving environment based on Unity (game engine) and Apollo (an open autonomous driving platform) as well as a real car. Unity provides ability to vary weather, road conditions and driving scenarios. Apollo includes all the code necessary for autonomous driving and consists of numerous modules: localization, perception, control, routing, safety module and others. This will allow the testing of algorithms in various conditions and scenarios in a safe way while being closer to real world. Possible challenges with realization of the HIL are discussed and highlighted in this paper.