The common goal of the project partners is to access various sensors for vehicle environment detection such as radar, camera and lidar via a powerful control unit. Actuators such as headlights and displays are also controlled by this unit. At the same time, the control unit evaluates the data from the environment sensors very quickly to create a virtual model of the vehicle’s environment in real time. This versatile functionality of the control unit forms the basis for various applications, such as driver assistance functions, spotlight for the targeted illumination of endangered objects, parking assistants or assistants for the protection of vulnerable road users. This innovative approach takes road safety and protection to a new level.
Robust sensor technology for reliable environment perception
The team of Christian Zinner from the AIT Center for Vision, Automation & Control is dedicated to the question of how intelligent vehicles can reliably perceive their surroundings even in difficult ambient conditions and poor visibility. The project uses modern 3D sensors, for which the AIT is researching suitable data processing algorithms. Among other things, fusion concepts for robust multimodal sensor configurations are also being investigated. The so-called imaging 3D radars and lidar sensors used here are currently being tested and will later become available for installation in series-produced vehicles. "Even today, there is still no 'ideal' 3D sensor that works equally well in all visibility and weather conditions while still providing an image of the surroundings in any detail. In addition to the 'fair weather’-properties of a sensor, greater interest must therefore be paid to its behavior in adverse conditions. This is where we bring in our many years of experience in the development of assistive and autonomous systems for a wide range of applications, where robust sensory perception of the 3D environment is a key component," says Christian Zinner.
Zonal architectures & control units
In order to be able to process this sensor data in the vehicle, the team at ZKW is significantly involved with the question of future EE architectures in the automotive industry and the associated ECU types. A clear trend is moving towards the so-called zone architecture and the zonal ECUs required for this. Various functions that were previously performed by individual ECUs will be bundled in defined zones. Here it is particularly important to cover the different developments and requirements of the market as best as possible into the course of the research work and to prepare for potential series developments.
The first demonstrator for this was already set up at the end of last year and is now gradually being put into operation and integrated into the zone architecture. To demonstrate the possibilities and functionalities, a first demo application for pedestrian detection targeting enhanced pedestrian protection is being implemented on the system.
Pedestrian detection using AI
The environmental data collected by the various sensors is processed using artificial intelligence (AI) methods in order to reliably detect particularly vulnerable road users such as children, pedestrians or cyclists. Training of the AI algorithms requires the analysis of thousands of different traffic situations. Therefore, a wide variety of scenarios are first simulated with the help of computer simulations and the sensor data required for the AI system is derived from this. The final fine-tuning of the evaluation procedures is carried out with the help of real measurement data recorded during test rides with the research vehicle. After completion of the training phase, different AI algorithms are available for the demo application for pedestrian protection. Since pedestrian detection function shall be integrated in the control unit running in real time, the focus of the development is on using methods that are particularly efficient in terms of computing power and energy consumption.
The zoneZ project has received funding from the Mobility of the Future programme. Mobility of the Future is a research, technology and innovation funding programme of the Republic of Austria, Ministry for Climate Action. The Austrian Research Promotion Agency (FFG) has been authorised for the programme management.
Please check also the presse release
Contact
Christoph Weidinger
Assistive & Autonomous Systems
Center for Vision, Automation & Control
christoph.weidinger@ait.ac.at
Christian Zinner
Assistive & Autonomous Systems
Center for Vision, Automation & Control
christian.zinner@ait.ac.at