The present study aims to add to the literature on driver workload prediction using machine learning methods. The main aim is to develop workload prediction on a multi-level basis, rather than a binary high/low distinction as often found in literature. The presented approach relies on measures that can be obtained unobtrusively in the driving environment with off-the-shelf sensors, and on machine learning methods that can be implemented in low-power embedded systems. Two simulator studies were performed, one inducing workload using realistic driving conditions, and one inducing workload with a relatively demanding lane-keeping task. Individual and group-based machine learning models were trained on both datasets and evaluated. For the group-based models the generalizing capability, that is the performance when predicting data from previously unseen individuals, was also assessed. Results show that multi-level workload prediction on the individual and group level works well, achieving high correct rates and accuracy scores. Generalizing between individuals proved difficult using realistic driving conditions but worked well in the highly demanding lane-keeping task. Reasons for this discrepancy are discussed as well as future research directions.

Modern computerized vehicles offer the possibility of changing vehicle parameters with the aim of creating a novel driving experience, such as an increased feeling of sportiness. For example, electric vehicles can be designed to provide an artificial sound, and the throttle mapping can be adjusted to give drivers the illusion that they are driving a sports vehicle (i.e., without altering the vehicle’s performance envelope). However, a fundamental safety-related question is how drivers perceive and respond to vehicle parameter adjustments. As of today, human-subject research on throttle mapping is unavailable, whereas research on sound enhancement is mostly conducted in listening rooms, which provides no insight into how drivers respond to the auditory cues. This study investigated how perceived sportiness and driving behavior are affected by adjustments in vehicle sound and throttle mapping. Through a within-subject simulator-based experiment, we investigated (1) Modified Throttle Mapping (MTM), (2) Artificial Engine Sound (AES) via a virtually elevated rpm, and (3) MTM and AES combined, relative to (4) a Baseline condition and (5) a Sports car that offered increased engine power. Results showed that, compared to Baseline, AES and MTM-AES increased perceived sportiness and yielded a lower speed variability in curves. Furthermore, MTM and MTM-AES caused higher vehicle acceleration than Baseline during the first second of driving away from a standstill. Mean speed and comfort ratings were unaffected by MTM and AES. The highest sportiness ratings and fastest driving speeds were obtained for the Sports car. In conclusion, the sound enhancement not only increased the perception of sportiness but also improved drivers’ speed control performance, suggesting that sound is used by drivers as functional feedback. The fact that MTM did not affect the mean driving speed indicates that drivers adapted their “gain” to the new throttle mapping and were not susceptible to risk compensation.

We conceptually replicated three highly cited experiments on speed adaptation, by measuring drivers' experienced risk (galvanic skin response; GSR), experienced task difficulty (self-reported task effort; SRTE) and safety margins (time-to-line-crossing; TLC) in a single experiment. The three measures were compared using a nonparametric index that captures the criteria of constancy during self-paced driving and sensitivity during forced-paced driving. In a driving simulator, 24 participants completed two forced-paced and one self-paced run. Each run held four different lane width conditions. Results showed that participants drove faster on wider lanes, thus confirming the expected speed adaptation. None of the three measures offered persuasive evidence for speed adaptation because they failed either the sensitivity criterion (GSR) or the constancy criterion (TLC, SRTE). An additional measure, steering reversal rate, outperformed the other three measures regarding sensitivity and constancy, prompting a further evaluation of the role of control activity in speed adaptation. Practitioner Summary: Results from a driving simulator experiment suggest that it is not experienced risk, experienced effort or safety margins that govern drivers' choice of speed. Rather, our findings suggest that steering reversal rate has an explanatory role in speed adaptation.