The majority of traffic accidents iscaused by driver errors. driver assistance systems are aimed to support the driver and can sustainably boost traffic safety and efficiency. The goal of the Institute of Transportation Systems is to examine driver behaviour, driver stress and strain, and accidents in order to derive the requirements for future driver assistance systems. The DLR combines its technological expertise with psychological and ergonomic research to produce effective driver assistance systems which satisfy the capabilities and expectations of the driver. Their implementation is evaluated during tests in a driving simulator, test drives and in real-world road traffic.
Knowledge of the status and behaviour of the driver, both with and without the aid of an assistance system, is crucial for a human-centered development of driver assistance systems. First of all it has to be understood how the driver masters the process of driving even in difficult situations, when his ability to work under stress reaches the limit and why driving errors occur. To achieve this, the Institute of Transportation Systems conducts empirical studies concentrating on driver errors and inadequate stress upon the driver. Psychological and physiological models of driver behaviour are developed, and the required functionality and intervention strategies for driver assistance systems are deducted..
Concepts and Technology
Driver assistance systems can only contribute to increasing traffic safety if they correspond with the needs of the individual driver. With this driver-focused approach, prototypes of assistance systems are developed which can be tailored according to the capabilities and expectations of each driver.
Steps taken by the Institute of Transportation Systems towards the successful implementation of an effective driver assistance system include carrying out empirical studies in order to determine significant assistance functions (e.g. lane change warning) and interaction strategies (e.g. acoustic warning signals.).
Driver assistance systems und functions are also outlined in their technical aspects. To achieve this, knowledge of the driver environment, information perception and information interpretation is required. Various technologies, including those from the aerospace industry, like image picture processing, positioning, communication, control- and planning algorithms are applied. In order to provide an effective, flexible method of integrating features, modular system architectures are developed. Research is also carried out on new and alternative human-machine interfaces.
Methods and Tools
In order to analyse driver behaviour, with and without the use of a driver assistance system, the Institute of Transportation Systems has developed a number of customised tools and test methods supported by its comprehensive infrastructure. With the test vehicle ViewCar analyses of driver behaviour in actual road traffic are carried out. With the driving simulator analyses of unusual or dangerous traffic situations are conducted under repeatable conditions. The Institute’s Virtual Reality-Lab and HMI-Lab allow for a flexible and economic testing of prototype systems without the implications of expensive hardware. The Institute’s dynamic driving simulator provides a realistic driving environment in whichprototype assistance systems can be tested, by combining a high-fidelity immersive visual system,with an integrated cockpit and a hydraulic motion system. The test vehicle FASCar, which can be controlled by a virtual co-pilot, is used for system evaluation purposes under real conditions and the SMPLab serves for quick testing of interaction strategies.