Research and /or Engineering Questions/Objective: In order to assess the effectiveness of today’s Advanced Driver Assistance Systems (ADAS) several consumer assessment tests deliver a platform with exact requirements, which need to be fulfilled to reach any rating. These tests currently include braking tests with tight boundary conditions and tolerances, which cannot be met by the inaccurate human driver. Therefore, the research objective is to set up an automated driving robot, which is capable of taking over the longitudinal and lateral guidance of the vehicle to assure a constant initial situation in which the emergency braking function needs to intervene. Methodology: The driving robot consists of a controller and steering, gas and brake actuators and is manufactured by Stähle GmbH. All actuators are put onto a seat-rail-mounting and all other hardware components are integrated onto one preassembled seat mounting, except for the robot controller. The set values are predefined in setups and the measured value is provided by an inertial measurement platform from GeneSys Elektronik GmbH. Using DGPS with correction data and the vehicles motion it delivers an accuracy of 1cm in the vehicles position. For enhanced data evaluation, a measurement system from Dewetron GmbH is used. It also works as the central windows-based interface for the user to operate the software of all components. At last, the Auto Mobil Forschung Dresden GmbH contributes to the partnership as an integrator and service provider that is responsible for test preparation, test execution, data evaluation and post processing. Results: The result is a complete system, which combines the hard- and software components seamlessly and allows an automated test execution for real world braking tests in any vehicle, including precise motion control and enhanced data acquisition. No changes to the car are required and no data like CAN from the vehicle itself is necessary to run the complete system. The software allows an individual adjustment of all predefined setups (e.g. for EuroNCAP) or to build a custom-made file. With this, the user is able to address all different kinds of driving scenarios with independent trajectories and driving programs. Next to safety tests, the system offers testing of vehicle dynamics, endurance testing and others, where strict testing requirements must be obeyed. Furthermore, an evaluation tool checks the boundary conditions and calculates the score for the executed test in order to receive liable data or to repeat a failed test. Limitations of this study: Today the target system for safety tests, which can be a Global Vehicle Target or a vulnerable road user, is not an integral part of the complete system. However, the communication to allow an interaction as a master slave system is already possible as well as data acquisition and evaluation. What does the paper offer that is new in the field in comparison to other works of the author: The main benefit lies in the automation and process-reliable seamless integration of all hard- and software components into a single complete system. The installation is easily done for almost any type of car and due to the predefined setups the application engineer does not need to configure or set up three different systems. Conclusion: Overall, the presented complete system is not a new methodology to allow a new test execution for new real world driving tests, but the complex parts needed to perform various development or assessment tests are now much easier to handle. Therefore, many users with less experience can operate the system and get answers to their research questions concerning the vehicles behavior or performance in various fields while minimizing error rates, costs and time required.
Tobias Rinnert, Auto Mobil Forschung Dresden GmbH, Germany; Günther Prokop