Project: Autonomous Vehicle for Automotive Testing And Research in Safety

On Automotive industry, the CO commitment must be to prevent accidents from occurring, and to reduce the importance of the driver's skills in favour of a better system performance. The last R&D project of this industry was fully focused on what is called “DAS”, Driver Assistance Systems._x000D__x000D_The active safety features ensure the smooth operation of the vehicle in motion; these provide stability, security and good response to driver commands. From this point of view, active safety systems should be understood as those devices intended to prevent an accident from occurring. _x000D__x000D_A rapid evolution in implementing these kind of systems during recent years has been experienced, having examples such as ABS, ESC, anti-collision systems, auto-braking systems, etc._x000D__x000D_During systems development, and in order to certify standard assessments, all those systems need to be tested on a safe environment and very repetitive conditions._x000D__x000D_The CO objective of this project is to Develop an unmanned mobile platform for testing those devices. _x000D__x000D_The platform will simulate a car, driving on a dangerous road in an urban scenario, running into the car carrying the active system to be tested. By using a simulator and a car, there is less human and material risk involved, and the repetitiveness and accuracy of the tests will be greatly increased._x000D__x000D_The Autonomous Vehicle for Automotive Testing And Research in Safety (AVATARES) will always reproduce a pre-programmed mission that will include a trajectory and speed profile relative to the vehicle under test, in order to assure the most accurate positions and speeds between the car and the platform, at the point of the hypothetical encounter. _x000D__x000D_This platform will carry on top many different soft dummies, such as a medium size car, pedestrian, bicycle or motorcycle, etc. in order to improve the accuracy of the weight and shape values of the simulated vehicle._x000D__x000D_Technologies_x000D_• Highly compact electrical mobile platform, based on electric car developments. _x000D_• Energy auto-generation or regeneration. Hybrid car technology_x000D_• Mechanical designs with new materials._x000D_• Low Level control algorithms for power train and guidance systems_x000D_• Autonomous guiding and synchronizing trajectories. Remote human operation._x000D_• Highly precise positioning system (global or dedicated) & Communications

Acronym	AVATARES (Reference Number: 6160)
Duration	01/07/2011 - 30/04/2013
Project Topic	Development of a autonomous vehicle that will be used to test safety devices for the Automotive Industry, saving material costs and reducing both material and human risks that can potentially occur during the testing operation through traditional methods.
Project Results (after finalisation)	Within the project AVATARES a new autonomous test system for the flexible and reproducible test of driver assistance systems was successfully developed by the Ps. In the part of the project conducted by Fraunhofer Gesellschaft (institute IOSB, application center AST), the CO focus was on the control of the mobile platform in terms of their navigation._x000D_In regard to this, firstly the necessary sensor equipment was identified, tested and adapted. Moreover, a localization system was developed. The basis of this localization system is a highly accurate RTK-GPS positioning system with position accuracy up to 1-2cm. However test drives with these system showed, that an autonomous control of a mobile platform at higher speeds, cannot be realized purely based on this system, due to effects of delay and sensor uncertainty, specifically in regard to the measured yaw angle. Therefore a data filtering algorithm was developed within the project on the basis of the concept of Extended Kalman-Filter. This data fusion approach combines the information about sensor noise and a system model to deliver an optimal estimation of system state, even for nonlinear systems, as it is for the mobile platform driving model. With this data filtering the relative accuracy of the position information could be improved significantly._x000D_Secondly a velocity-adaptive, nonlinear track controller was specially developed. This track controller was tuned in simulation and later in final field tests in Madrid and delivered very good results in lane keeping. This lane keeping behavior of the platform is the basis for all implemented maneuvers of the test system._x000D_Thirdly the overall control system was developed and realized by a real-time capable control unit. To achieve this in project time, the approach of rapid-prototyping by using xPC Target Toolbox (Matlab/Simulink (R)) was followed. With this an independent control system was developed, which allows the control of a variety of test maneuvers including controlled acceleration and deceleration, roundabout, lane change and path invasion, which were defined in the specification work package of the project. The control system manages the information flow and triggers the track controller to achieve the maneuver control of the platform, even in interaction with the vehicle under test (VUT)._x000D_Fourthly a mission planning software (AVATARES mission planner) was developed. The software allows the creation of test maneuvers, the transmission of these maneuvers to the control system and the monitoring of the control system and the mobile platform during maneuver execution_x000D_
Network	Eurostars
Call	Eurostars Cut-Off 5

Project partner

Number	Name	Role	Country
2	ARIES INGENIERÍA Y SISTEMAS, S.A.	Coordinator	Spain
2	Fraunhofer Gesellschaft zur Förderung der Angewandten Forschung e. V.	Partner	Germany