Kempten University of Applied Sciences adopts new simulator to investigate autonomous driving technologies
Hochschule Kempten’s Adrive Living Lab in Germany will adopt an advanced vehicle driving simulator (aVDS) from automotive test system supplier, AB Dynamics. The new simulator is expected to enable the university to conduct thorough research into driver interaction with advanced driver-assistance systems (ADAS) and automated driving (AD) systems. The aVDS is a third-generation vehicle dynamics-grade driving simulator and has been designed to provide the automotive industry with unprecedented research and development capabilities for autonomous systems. Installation will be completed by the second quarter of 2019.
The Adrive Living Lab is focused on the development of autonomous driving, with an emphasis on the driver’s interaction with autonomous systems. It investigates the effect of these systems on drivers’ perceived safety and comfort, the two parameters often identified as the most important for consumer acceptance and enthusiasm. The level of demand for such systems stems directly from the views of the drivers that use or interact with them. The new aVDS will allow researchers to put a driver in the loop and gather both subjective and objective feedback on the performance of current and future driver assistance and autonomous systems. It will also allow the university to investigate how to effectively use a simulation-based method and toolchain for evaluating these technologies.
“Driver-in-the-Loop (DiL) simulators are essential to assessing the capability of ADAS and AD at an early development stage,” explained Professor Bernhard Schick, Kempten University of Applied Sciences. “Virtualization is an important trend to increase the effectiveness and efficiency of ADAS and AD development. Real prototypes are expensive and becoming less common. Complex vehicle test scenarios are difficult to carry out on real roads. For engineers, however, it is very important to be able to experience and evaluate new functions often, even without a physical prototype available. Researching and understanding how normal drivers experience and interact with the new features is also very crucial to creating best-in-class functionality and usability. A key benefit of a simulator like the aVDS is that we can explore the subjective as well as the objective effects to the driver under a variety of circumstances.”
Hochschule Kempten intends to use its aVDS to measure drivers’ stress levels and investigate the safest ways to conduct a handover—when control is returned to the driver—if the system is not capable of performing a maneuver.
“We are convinced that the aVDS from AB Dynamics is the best available solution currently on the market for our key applications,” continued Schick. “The advantages afforded by high dynamic capability and low latency response times are significant. The driver’s input can be transferred into vehicle motion without excessive time delays, making the experience precise and therefore realistic. This is especially important for the vehicle motion outside of the center of the motion envelope. We were also convinced by the large motion travel of the platform, which is effectively extended by advanced cueing. In addition, the aVDS can be installed much more easily within the confines of existing infrastructure, which will save us both time and money.”
The university also intends to capitalize on the simulator’s architecture by mounting a real car steering system to the platform, from the steering wheel to the ball joints. A steering rack can be difficult to model because it has many nonlinear components, and there are significant advantages to including it within the simulator’s feedback loop. By using the simulator’s hardware-in-the-loop capability, the team at Kempten will avoid the inherent inaccuracy of a modeled system. Due to the simulator’s kinematics, stiffness, and low center of gravity, the feedback and steering feel that drivers receive is incredibly realistic. With this level of fidelity, the aVDS can then be used to perform a variety of steering-based testing, including the evaluation of on-center-feel and driver response.