This guest contribution to Automoblog is courtesy of Paul Zubrinich, Chief Marketing Officer for Optalert, an Australian company that specializes in detecting drowsiness and other cognitive states in eyelid movements. In addition to supporting automakers as they improve their drowsiness detection systems in new vehicles, Optalert discovered the first non-invasive method to screen for obstructive sleep apnoea while the test subject is awake and is making progress in early screening of Alzheimer’s, Parkinson’s, and other neurodegenerative conditions.
If you’re in the market for a new car, you have likely heard about driver monitoring systems (DMS). A DMS uses sensors and software to monitor the driver’s behavior and condition, helping to improve safety and performance. Some of these features are mature and reliable, while others are quite hit-and-miss.
Why Do Vehicles Have a DMS?
There are three broad drivers pushing driver monitoring systems into the market:
First, new regulation, especially in Europe, has mandated that cars must detect inattention, drowsiness, and other forms of driver impairment. There are also talks at the federal level in the United States of regulation around intoxication detection. It appears to be a growing frontier in safety regulation in automotive.
Second, autonomous vehicles are encountering numerous technical challenges and will likely take a lot longer to get right than many technology optimists initially predicted. As cars continue to be operated by humans, monitoring the driver’s state will make them safer.
Third, a lot of advanced driver-assistance system (ADAS) features have annoyed drivers in recent years. For example, a colleague regularly tells me of an annoying lane-keeping assistance feature in his car that invasively pulls him back into the lane when he is consciously and purposefully changing lanes.
A catchphrase in the automotive industry is “DMS will save ADAS.” If a DMS could see he was perfectly alert and intentionally deciding to do something, the ADAS system could leave him alone. But if he was hurtling along at high speed in the rain, sleep-deprived, and visibly distracted looking at his phone, the ADAS could intervene more based on input from the DMS.
Driver Monitoring Features: The Good, The Bad, The Clumsy
To some drivers, it might feel a bit paternalistic and restrictive to have software monitoring you as you drive. If this sounds like you, consider selecting a new vehicle with a DMS that has the option to disable certain features. Many automakers include a “Reduced alarm mode.” I’ve seen under the hood of many technologies at various trade shows and automotive industry events. Engineers are still ironing out a lot of kinks with these systems, but I’m confident that software updates and product recalls due to annoying false alerts will become more common in the next few years.
When you take a car for a test drive, ensure you also scrutinize the DMS. Does it beep unnecessarily? Can you turn off various alerts? Would a minor bother now grow into a pet peeve if you were subjected to it every time you drive? Don’t take at face value what it says in the pamphlet or reassurances from the dealer. Test it for yourself. Trust your experience.
With that in mind, here are the top five DMS features you should evaluate if you are considering a car with the technology.
#1: Driver Identification
Driver identification uses facial recognition technology to identify the driver and adjust the vehicle’s settings accordingly. It can remember multiple drivers’ preferences, including seat position, mirror angles, and even preferred temperature. This personalized experience saves time and enhances comfort.
Most DMS systems have it, and it is usually a very reliable technology, although some facial recognition algorithms have had issues with certain ethnicities due to limited training data, although this is becoming less of a problem with the emergence of synthetic data.
#2: Distraction Detection
Distraction is a prominent cause of accidents on the road. A good DMS should be able to detect when the driver is distracted, whether it’s by a phone, the radio, or something outside the vehicle. The system can then alert the driver or take corrective action to prevent potential accidents. Most systems do this with a camera that detects the direction of the driver’s gaze and sometimes objects such as phones. This type of DMS system can also provide parents of teen drivers additional peace of mind.
One thing to note is how some systems might be too sensitive and prone to sounding bothersome alerts when the driver is simply checking their blind spot or glancing at their speedometer. During your test drive, check if it beeps unnecessarily when you take your eyes off the road for a legitimate reason. That will confirm if it triggers false alerts. Try not to “over-test” the system, however, as taking your eyes off the road for an extended period of time can be dangerous.
#3: Drowsiness Detection
Drowsiness is one of the most common causes of road accidents and, thus, a vital DMS feature. Its job is to detect when the driver is dangerously drowsy, then sound an alert or even take control of the vehicle to prevent an accident. Algorithms to detect drowsiness vary significantly in how they work and how well they work. A DMS may use the driver’s eye or eyelid movements, facial expressions, or other physical signs.
Drowsiness is something that many engineering teams have struggled to measure accurately. A rigorous test of a drowsiness detection system requires trained experts in sleep science, so this isn’t something you can fully evaluate yourself on a test drive. However, there are some simple things you can check for.
First, check if it responds to yawns. If so, it is an indicator of a poor system, as yawns are not always related to drowsiness. Next, play or act being tired in a very performative way. If it responds to this, that’s also a red flag. When we dramatically pretend to be sleepy, it is far from what a drowsy person really looks like.
In short, if you are not truly sleep-deprived to the point of overwhelming tiredness, you will not be able to trigger an alert in an accurate drowsiness detection system. If you are not sleep-deprived and can trigger a warning, you can be certain the system will annoy you with false alerts as you drive.
#4: Specific Activity Detection
Specific activity detection is a more advanced feature found in some DMS systems. It detects particular activities like drinking, eating, texting while driving, or not wearing a seatbelt. By detecting these activities, the system can alert drivers and encourage safer habits behind the wheel.
#5: Child Presence Detection
Child presence detection is required for cars to get the highest safety rating in Australia and Europe. It uses various sensors to detect if a child has been left in the vehicle, alerting the driver or even authorities if necessary. To test for false alerts, try using a doll or stuffed animal. Advanced systems can differentiate between a person that is breathing and a childlike object that is inanimate.
Latest DMS Features Summary
More vehicles are bound to appear on the market with DMS technology. If and when you are in the market for a new car, knowing a little something about the five features above will help. DMS systems can significantly enhance safety and provide a more personalized driving experience. If you test the car for yourself and keep an eye out for false alerts, you should be able to determine if it suits your needs and does what it says in the brochure. Happy shopping!
Paul Zubrinich is the Chief Marketing Officer for Optalert, an Australian company that specializes in detecting drowsiness and other cognitive states in eyelid movements. Optalert has been measuring drowsiness in drivers for over 15 years and is actively supporting automakers to improve their drowsiness detection systems. Optalert’s technologies are based on blepharometry, a term introduced by founder Dr. Murray Johns to define the unique study of measuring eyelid movement captured through a sensor or camera. Outside of automotive, Optalert discovered the first non-invasive method to screen for obstructive sleep apnoea while the test subject is awake and is making progress in early screening of Alzheimer’s, Parkinson’s, and other neurodegenerative conditions.