As vehicle automation becomes a reality we will see the nature of the driver-vehicle interaction begin to change, taking on a more supervisory role with higher levels. It is, however, not unlikely that driver's role will switch from time to time from that of a passive supervisor to an active agent, requiring the driver's attention and/or intervention.
Rightly, there are two consequences of this that are currently receiving much interest in the human factors community: 1) assessing the driver's state during automation such that we might understand their readiness to resume control, and 2) assessing the capabilities and limitations of the driver once they have taken back control. While reaction time has been the gold standard for performance, in the context of vehicle automation faster reaction time may not necessarily be the desired outcome: the scenarios that would see a driver have to take back control will likely not only require control-level responses, which means that if the driver needs to make a tactical decision first then asking how long it takes the driver to take control is less important than the quality of the decision they make once they do.
My questions:
1) Given that a driver's response will be linked to the design of the actual system, and few systems currently exist, how do we design the system and scenarios in our experiments such they are meaningful to us and vehicle manufacturers?
2) Without vehicle-based metrics, how does one go about assessing driver state and readiness to intervene/take over/make decisions?
3) What are the pros and cons of using such measures as eye-tracking (which aspects?), fNIRs, EEG, HRV, and other psychophysiological measures to assessing readiness? Are there others?
4) How might one go about assessing the quality of the transition?
5) How can we link the various driver state and performance-based metrics to build a meaningful assessment of the driver's capabilities and limitations in the transitions?
This is not at all my specialist area, but it occurs to me that 1) and 5) are particularly important in relation to answering 4). The key contribution that you could be seeking to make is in providing evidence that will shape the direction and outcome of the transition to road transport automation.
This is not at all my specialist area, but it occurs to me that 1) and 5) are particularly important in relation to answering 4). The key contribution that you could be seeking to make is in providing evidence that will shape the direction and outcome of the transition to road transport automation.
This is not at all my specialist area, but it occurs to me that 1) and 5) are particularly important in relation to answering 4). The key contribution that you could be seeking to make is in providing evidence that will shape the direction and outcome of the transition to road transport automation.
This is not at all my specialist area, but it occurs to me that 1) and 5) are particularly important in relation to answering 4). The key contribution that you could be seeking to make is in providing evidence that will shape the direction and outcome of the transition to road transport automation.
One upcoming technology for monitoring the driver's alertness and other ergonomic factors would be through electrophysiological signals such as EEG or EOG. Essentially, this might be uncomfortable on a rudimentary scale due to involvement of electrodes, etc which might make the subject uncomfortable. But given the scale of development in better and user-friendly acquisition systems like Mindwave EEG headset, etc. with wireless capabilities this would in coming days not pose a problem.
Another major area would be through the visual information of the driver through a camera, as many people are doing so. But, this again faces problems in concern of minimum ambient light, etc. But by this concept of body based signals, we can characterize blink-rate variability to asses factors like blink-to blink duration, eye opening time, closing time, etc. All of such factors can be quantified to assess the driver's state. Extended analysis can be done even on vigilance monitoring to gauge vigilance index, etc.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Human-Computer Interaction