The field study described in this report in a replication and extension of the study that concluded the DRIVE I project V1041 GIDS in 1992. The present study was made possible because of the availability within the ARL4DNE consortium of a range measuring radar. It took 45 driver-subjects out on motorways and other thoroughgoing roads under three conditions: (1) A control condition; (2) A non-integrated condition, in which the different support components (navigation; and an active accelerator acting as both a collision avoidance and speed-limiting device) worked without taken into account each other's existence; and (3) An integrated condition, in which messages from the different support components were scheduled according to their priority. Driving performance was measured in terms of navigation errors, and of (distributions of) vehicle speed, acceleration, deceleration, and time headway to preceding vehicles. Driver workload was assessed in terms of subjective ratings, the 0.1 Hz component of the heart rate spectrum, and performance on a secondary task. Beneficial effects of support were apparent in several performance variables, notably the distribution of vehicle speed and of time headway. With respect to the latter it turned out that the collision avoidance system almost completely eliminated the occurrence of short (<1.0s) and very short (<0.5a) headways. The workload data showed, in the subjective ratings, that the non-integrated condition can generate more workload than when the different support components act interrelatedly. This pattern was, however, not apparent in the heart rate and the secondary task data. Reasons for this are probably in the different aspects of workload that these parameters index. The report concludes by stating that the beneficial effects of GIDS support obtained have to be checked in long-term experimentation, that is, by following driver-subjects as they gain experience with the support. (A)
Samenvatting