Countermeasures

In the last episode, we discussed how an objective drowsiness measure is implemented in the field. We also briefly touched on techniques called countermeasures that can temporarily alleviate drowsiness. In this episode, we first talk about the way drowsiness systems are built today and their pitfalls. We then define countermeasures and discuss how they can be used to extend safe drive time.

What are countermeasures? Countermeasures, also known as interventions (Euro NCAP) are actions that can be undertaken by a person or programmed into a car’s Advanced Driver-Assistance Systems (ADAS) to prevent further harm to the driver and other road users. Countermeasures are ranked based on their intensity. Early stages can relieve a person’s drowsiness levels by providing audio/visual indicators that a person is drowsy, whereas end stages are intended for unresponsive drivers where an automated safety manoeuvre (also known as the Minimal Risk Manoeuvre or MRM) is executed to ensure the safety of all road users.

With modern advancements in Advanced Driver-Assistance Systems (ADAS), cars can now be programmed to provide a multitude of countermeasures to a driver when necessary. However, in their present forms, many drowsiness detection systems, regulations and assessment programs classify drowsiness as a simply defined yes or no rather than a state of continuum. As and when the Driver Monitoring System (DMS) indicates that a driver is drowsy, the car would have to produce all countermeasures. This could have adverse effects:

1. If a driver experiencing early-stage drowsiness is served with all available countermeasures, it would annoy the driver, leading them to shut these safety systems off.
2. If a car only produces warnings when a driver is already experiencing late-stage drowsiness, microsleeps or sleeps, the driver might be too far down the path for the warnings to be effective.

These increasingly complex ADAS systems need an effective strategy for delivering warnings and countermeasures to drivers, and drowsiness levels can form the basis for when these measures should occur. Implementing countermeasures that are proportional in severity will not only ensure drive comfort and quality without compromising on safety, but can also be used to extend safe driving duration, as seen in our commercial fleet example. However, countermeasures work best if a person’s drowsiness level can be objectively and accurately quantified on a graduated scale, providing the resolution necessary to differentiate between early-stage drowsiness, late-stage drowsiness and everything in between. Pairing different levels of drowsiness with their appropriate countermeasures provides us with a system that can prolong safe driving time, ensuring that we arrive at our destination safely.

Once a driver’s drowsiness level is quantified, a multi-level countermeasure approach to remedy their drowsiness is as below:

1. Awake: No special countermeasure required from the system. All ADAS warning systems are set to the lowest level of sensitivity, giving the driver liberty to drive in comfort and at ease.
2. Early drowsiness onset: Preservation phase – to allow the driver to focus on the main driving tasks, the system “makes driving easy” by enabling automatic cruise control, lane maintenance, and reducing distractions through warning systems.
3. Early-stage drowsiness: Driver stimulation phase – the system aims to stimulate the driver to counter further drowsiness by enhancing feedback about lane keeping accuracy, applying acoustic and haptic feedback about lane departure, and amplifying kinesthetics feedback of the adaptive cruise control (stronger acceleration/deceleration).
4. Drowsiness: Modification of driving conditions – the system aims to introduce sensory stimulation to the driver by modifying the ambient conditions: introducing fresh air (ventilation or opening window slightly), increasing radio/entertainment system volume, and reducing cabin temperature.
5. Late-stage drowsiness: High stimulation phase – the system aims to increase driver engagement, “making driving difficult” to further engage the driver in the driving activities by disabling automatic cruise control, increase steering torque (resistance), reducing spring suspension, amplifying motor sound, and even introducing controlled disturbances.
6. Falling asleep: Protection phase – the system tries to compensate for the driver’s potential impairment by advising the driver to stop and removing limitations placed on all warnings.
7. Sleep: Unresponsive phase – introduction of MRM and stopping the car safely.

This approach means that drivers will only receive the countermeasures that are appropriate to their current state. Drivers experiencing early-stage drowsiness will receive more subtle countermeasures that will temporarily reduce their drowsiness level and extend safe driving time without overwhelming them with other countermeasures. Drivers experiencing mid- to late-stage drowsiness will receive more intense countermeasures to further reduce their drowsiness level, inform them of their current state and suggest that they stop and take a break whenever it is safe to do so. Drivers who are falling or who have fallen asleep will receive the harshest countermeasures to wake them up, or in the worst-case scenario, automatically pull the car over in the safest way possible. This strategy ensures that drivers are not disturbed when they are not at risk of drowsiness-related performance impairment but, when necessary, introduces countermeasures that are appropriate to the driver’s current drowsiness level.

An example of this is depicted in the figure below. Because we introduce different levels of countermeasure intensity based on the driver’s drowsiness severity (orange), each level alleviates drowsiness by a small but significant amount. Even though over time a driver’s drowsiness does keep accumulating, the compounding effect of implementing a multi-level countermeasure strategy results in an extended safe driving duration, delaying the time it takes for a driver to enter a state where they are at high risk of an accident. In this example, the multi-level countermeasure strategy successfully prevented the driver from being at high risk of an accident through the course of their drive. In addition, because each level delivers a different set of countermeasures based on its intensity, the driver does not feel annoyed by or become desensitised to the alarms. This is in stark contrast to a single level countermeasure strategy where a driver’s drowsiness level is defined as a binary outcome (blue). Because all countermeasures are delivered at a single point (be it too early, on time or too late on the drowsiness scale), the reduction in drowsiness level only occurs once or twice, and thus we do not see the compounding effect when compared to the multi-level countermeasure strategy. Consequently, drivers may become too annoyed by the alarms (and thus decide to disable them) or may become desensitised to these repetitive alarms (red marker). This inevitably results in a continued increase in drowsiness levels, placing the driver at a high accident risk for the remainder of the drive (grey).

By coupling an objective drowsiness measure with a multi-level countermeasure strategy, our system aims to strike a perfect balance between drive comfort and safety, moving us closer to a world where drowsiness-related accidents can be a thing of the past.

And that’s a wrap! We hope that you have enjoyed this drowsiness mini-series. Over the last ten weeks, we started by discussing what drowsiness is and how our body regulates it. We then talked about how drowsiness is quantified and why objective measurements are important when determining how drowsy a person is. The story then moved onto insights that we have gained from drowsiness detection systems currently implemented in commercial fleets, and finished off here with countermeasures as a tool to balance how we deliver drowsiness alerts (therefore ensuring a positive driving experience) and safety (by extending safe driving time).

THE END… roll credits

Over the course of this mini-series, we have received numerous drowsiness-related questions. We love the enthusiasm that you’ve shown about drowsiness, so we have decided to compile and address them in an epilogue! In the meantime, for those who have more questions (but didn’t know where to ask them), please send us an email at info@optalert.com. If you are developing a DMS system and want to incorporate Optalert’s advanced drowsiness monitoring, release 7.0 of our software development kit has just been announced, and will be available from 30 June 2022. For more information please email us at automotive@optalert.com, or download the release notes. 

Thank you for joining us, stay safe!

Drowsiness Detection and Validation

Assess your Driver Monitoring Solutions against Optalert’s scientifically validated measures.

Upload your dat