IDA Technical Reports: abstract

Malec, J., Morin, M., and Palmqvist, U. (1994). Driver Support in Intelligent Autonomous Cruise Control. Technical Report LiTH-IDA-R-94-26, Department of Computer and Information Science, Linköping University, Sweden. Accepted to Intelligent Vehicles Symposium'94, Paris, France, October 24-26, 1994. (bibtex),

Abstract: The next-generation cars will be furnished with a multitude of advanced electronic equipment to enhance safety, increase driving comfort, improve economy and decrease environmental effects. Examples of devices to this end are autonomous intelligent cruise control (AICC), route guidance systems, and road/tyre friction estimation systems. Because these support systems generate a lot of data, the amount of information potentially available to the driver is very large, and sometimes contains even conflicting messages. In such situation there is an immanent danger of overloading the driver's perceptual and reactive capabilities. Clearly, there is a strong need for systematic handling of the output to the driver from different subsystems in the car, in order to guarantee that the driver gets the relevant information at the right time. This task is complicated by a number of circumstances, for instance: 1. Output requests may come from many different sources (support systems) operating independently, or almost independently, without a coherent view of the information available in the whole system. 2. Information is time-dependent, so it might prove useless, or even dangerous, if presented to the driver at the wrong time. 3. The amount of information the driver can handle is not untimed nor static, but varies dynamically depending on, among other things, the traffic situation, the weather conditions and the kind of manoeuvre the driver is performing. 4. There may be several output channels available (for example displays, voice output, and haptic actuators) whose appropriateness may change dynamically. 5. Drivers may prefer some output channels to others, or may want to suppress some kinds of messages (e. g. speed excess messages). In this paper we describe a tentative design of the driver support system to be integrated with the Volvo demonstrator car for the AICC function. The design is based on the concept of a driver information unit (DIU) which is able to represent both the dynamics of the environment and the driver's workload and preferences. The purpose of the DIU is to arbitrate between different communication requests on the basis of the current driving situation and to present the information to the driver using the most appropriate medium. The DIU is also able to filter the information according to the driver's expressed preferences. So far, the DIU builds and maintains a limited model of the environment from available sensor data and output from the support systems. It does not, however, predict or plan in advance what might happen, but the system design permits such services to be added, if required.