Fatigue Monitoring System Article

Overdrive
November 2, 2017
Wearable fatigue monitoring devices competing with camera systems
By Max Heine

Fatigue monitoring technology has been used for decades, though until recently only on a very limited basis in trucking.

The technology made its first major inroads in mining, where closing a mine for a day or more after a fatal accident can cost the owner millions of dollars. For that reason, the return on investment and the willingness to try a new approach to safety made a lot of sense, says Daniel Bongers, chief technology officer for SmartCap Technologies, the company that evolved out of his early work in fatigue research and mine safety.

SmartCap and at least two of its fatigue-monitoring competitors, Seeing Machines and Optalert, have global operations based in Australia, where mining of coal and other minerals represents a huge part of the country’s economy.

The initial focus on mining opportunities is one reason fatigue-monitoring vendors have achieved so little penetration in the U.S. trucking market, Bongers says.

That’s changing fast. Fleets representing tens of thousands of trucks use some form of camera systems for safety reasons. Those reasons increasingly include fatigue monitoring, especially as driver-facing cams get put into cabs. While adoption of wearable fatigue monitoring systems lags the adoption of camera systems, wearables are making headway as more fleets test the products.

Two notable differences some wearables vendors claim are that their systems, relative to driver-facing cameras, are less invasive and more suited to predicting fatigue. Some not only rate a driver’s alertness before he starts to drive but also promise to predict to the hour when the driver will cross a fleet-chosen safety threshold.

“I see a lot of systems come and go,” says Richard Hanowski, director of the Center for Truck & Bus Safety at the Virginia Tech Transportation Institute. Many systems based on cameras or other technology look good during testing but fall short in real-world applications for various reasons, including driver resistance.

Driver-facing cameras in particular are “a very strong invasion of privacy,” says Avi Geller, CEO of Maven Machines, whose headset measures head movements to detect driver fatigue or distraction. “We hear it all the time, ‘The truck is my home. I live there. You want me to put a camera in my home … It’s disrespectful.’”

“I could abide a device that detects fatigue,” says Dave Reid, a company driver for Pride Transport, “but I could not abide any video or audio recording in the cab and wouldn’t work with a company that mandated that.”

“Driver fatigue? How about just asking?” says Michael Wright, an owner-operator leased to Laser Transport. “Or trust the driver’s word that they are tired and need to take a nap before continuing on? My God, the more technology, the stronger the ‘stupid’ grows in this industry.”

Driver objections to fatigue monitoring, especially driver-facing cameras, aren’t the only problem in these emerging technologies. One is that planning drivers’ schedules to consider all the predictive and evolving real-time data supplied by these systems could pose big challenges to a fleet’s ability to manage its own logistics and maintain on-time deliveries.

That poses a “major problem” industrywide, says John Hopkins, a company driver for XPO Logistics. “Can you see Fed Ex, XPO or UPS shutting down when it has to be there overnight? … I suspect we’ll just turn a blind eye for years to come. The industry and government doesn’t really want to know.”

Another problem is that even when a system delivers an accurate reading that a driver is fatigued, that warning doesn’t fully address the need to get the driver off the highway quickly for rest, Hanowski says. That’s one of “a lot of pitfalls on the way” toward developing a fully accurate and practical system. Nevertheless, “there’s a lot going on in this area,” including some leading technologies that are collaborating with his center.

How wearable devices detect fatigue without cameras
By Max Heine

One method of fatigue monitoring involves measuring a driver’s head motion without using a camera. Maven Machines’ Co-Pilot headset “detects if a driver is looking forward through the windshield or looking up or down or sideways,” says Avi Geller, CEO.

Fatigue-monitoring technology has made its largest inroads in trucking through road-facing cameras, and more recently through dual-cam systems that include a driver-facing cam. Truck-mounted video has been available for years, but the emergence of vendors who help analyze the footage for driver coaching, legal protection and fatigue monitoring has expanded the value of these systems.

Relatively new to trucking are wearable devices that detect fatigue without requiring camera input. At least two of the wearables depend largely on actigraphy, a decades-old science. Typically worn like a wristwatch, an actigraph measures body movement to assess sleep debt.

The sleep history records are combined with other data through each company’s own algorithm. It’s mainly the actigraph’s cumulative records of sleep quantity and quality over recent days that, with other factors, enables the complete system to forecast how quickly alertness will decline.

Another decades-old technology, electroencephalography, is the foundation for the SmartCap Technologies system. EEG, which measures brain waves, has long been acknowledged as the “gold standard” for fatigue detection, says Daniel Bongers, SmartCap’s chief technology officer.

Following are some of the leading monitoring systems that do not require cameras

Fatigue Science’s Readiband actigraph

Fatigue Science is testing its Readiband actigraph with fleets, says Jacob Fiedler, the company’s sales director. The system was developed by the U.S. Army Research Lab with funding from the U.S. Department of Transportation.

“Our core value proposition is we can give you the ability to see at what point you’ll be fatigue-impaired hours in advance,” Fiedler says. Other systems might do a good job of detecting fatigue, “but it’s not addressing people’s sleep patterns, informing the individual how to improve upon the individual’s sleep. The only cure for fatigue is sleep.”

The system’s software produces a driver score that starts with the current level of alertness and predicts its rate of decay up to 20 hours. If the driver takes a nap, it will boost his immediate and projected scores.

Such emphasis on remedial approaches in light of projected fatigue is a big part of the sell for predictive systems, whose vendors argue that camera-based systems are of limited use after delivering warnings to a driver and dispatcher.

Some of Fatigue Science’s U.S. trucking clients “have asked us to explore the association in their fatigue data and their telematics,” Fiedler says. If there is an association between fatigue and performance such as hard braking and speeding, that means there is potential not just for safety but also for cost-cutting.

Fatigue Science is working with a trucking fleet, which Fiedler declined to name, that has medical personnel on staff and is using its data to work with drivers. “It’s not a diagnostic tool,” he says, but “it’s certainly an awesome pre-screening tool.”

In some cases it can point toward testing for obstructive sleep apnea or another chronic condition. In others, it can be a matter of coaching on the benefits of napping or other fatigue mitigation.

Fatigue Science has worked with not only trucking fleets but also customers in mining and construction, as well as the military and professional sports. The trucking fleets include the Canadian company Arrow Transportation. Fiedler declined to name other fleets, one of them a “global operator,” because they are still in testing stages.

Caterpillar is a “global delivery partner” with Fatigue Science, primarily in mining and construction. Cat uses its own brand on the Readiband actigraph and also has worked with SmartCap’s fatigue monitor that measures brain waves.

Curaegis actigraph system

One of the monitoring products that relies on actigraphy is the CURA (Circadian User Risk Assessment) system from Curaegis.

“Some people degrade slowly” in their level of alertness, says Chief Executive Officer Richard Kaplan. “Everybody’s different.” After detailed hourly lab tests were compared with the CURA system’s predictions, “it’s proven we can give real-time fatigue levels with laboratory accuracy,” he says.

Curaegis recently finished free beta-testing with clients from its target markets of trucking, busing and other fields. Paid pilot testing was to follow, with clients that include about five truck fleets, Kaplan says.

The company is in talks with six or seven insurers about the potential for its myCadian actigraph device to reduce accidents. “Some have told us if we can cut accidents 1 percent, it would mean a lot,” says Kaplan, who believes the reduction could be as much as 50 percent.

Curaegis also is working on compatibility with Apple watches that would allow the watch to function not only as an actigraph but also capture data that its myCadian device uses, such as body temperature, heart rate and blood oxygen level.

Curaegis, like some of the other fatigue-monitoring vendors, stresses training and actions to be taken as a driver becomes fatigued. Curaegis’ Z-Coach intervention product has an online course that teaches users about sleep and fatigue mitigations to boost a driver’s fatigue rating. A 20-minute nap for someone at the danger threshold often can boost alertness to a safe zone that lasts for four or five hours, Kaplan says.

Z-Coach also highlights the need for potential interventions when fatigue might be indicative of medical issues other than an ordinary lack of sleep. Doctors helped prepare a screening in which drivers answer questions that could lead to further diagnosis for sleep apnea, insomnia, low blood iron content or other problems.

Maven’s Co-Pilot headset

Another approach to fatigue management involves measuring a driver’s head motion without using a camera. Maven Machines’ Co-Pilot headset “detects if a driver is looking forward through the windshield or looking up or down or sideways,” says Avi Geller, CEO.

The key measurement is mirror checks, which should occur every five to eight seconds, according to the Federal Motor Carrier Safety Administration, Geller says. “As people get more fatigued, their mirror check rate decreases.”

The headset also detects head bobs and jerk-up motions, both signs of microsleep.

“The data does show monitoring this kind of head motion does indicate fatigue,” he says. The headset also detects head movement that suggests distraction, such as looking down to read a phone screen.

The Maven system integrates head motion readings with data from other sources such as weather, traffic and the truck’s electronic control module.

Using Maven headsets for the last year, in combination with safety training based on a road-facing camera system, enabled Conard Logistics to go a year without accidents, says co-owner James Griffith. The 35- to 40-truck fleet is based just outside of Nashville, Tennessee.

“I think it’s a really up-and-coming system that’s going to be real popular,” Griffith says. “We test anything if we think it’s going to help with our safety.” His drivers have not complained about the system and like it because it’s easy to use, he says.

Geller says Maven has been engaged with a partner, which he declined to name, to possibly integrate road-facing cameras to increase the system’s accuracy.

There’s also potential for integrating performance data other than camera input. Linking hard-braking events with fatigue indicators can yield highly accurate results for subsequent fatigue intervention, Geller says.

In 2015, the company signed PGT Trucking, based in Monaca, Pennsylvania, as its first U.S. trucking fleet customer. More than 15 trucking fleets now are testing the Maven system, Geller says.

“The scientific concept is valid,” says company driver Bob Stanton, who has tested the system. His involvement with sleep apnea problems has led to further involvement with fatigue specialists.

Stanton notes that when he uses lively music to fight fatigue, he tends to bounce in his seat, but the system interprets the motion as a mirror check.

His experience isn’t really misleading in that “he is being active, not microsleeping,” Geller says. “He is trying to keep himself alert and focused,” sort of “tricking the system to detect the right thing.”

SmartCap’s LifeBand

SmartCap Technologies uses a head-worn device that detects microsleeps before the visible fatigue signs of long eyelid closure or head nods. Its Life by SmartCap system uses electroencephalography, a 40-year-old science that about 20 years ago became recognized as the only way to discern the true onset of microsleep, says Daniel Bongers, chief technology officer.

More than half of a microsleep takes place before eye closure or head nods, he says. “That person has already been asleep two to 12 seconds, certainly more than enough time to have driven 400 yards,” Bongers says.

Whereas the makers of actigraph-based systems tout their ability to predict a driver’s rate of decline in alertness for up to 20 hours, SmartCap’s system is mostly a real-time indicator. However, its narrow window of what Bongers says is a more refined prediction is crucial, he argues.

That warning of being “at risk of becoming at risk,” he says, allows the driver more leeway in mitigating what can be a short but dangerous transition toward sleep. It allows for simple actions such as sipping a drink or sitting up. If the warnings persist, the driver needs a more vigorous response, such as pulling over for a brief walk or other exercise, which can raise the odds of safely reaching an appropriate place to nap.

“Humans are complex,” Bongers says, noting that there are notable differences in fatigue between men and women, as well as the young and old. Also, research shows there are 243 “physiologies” that indicate distinctions in sleep cycles and alertness, which makes it difficult to make meaningful forecasts of alertness for all drivers, Bongers says. “There’s very little connection between activity tracking [from an actigraph] and its relationship to fatigue.”

Following the company’s early market efforts in mining and construction, “We’ve only recently put a focus on over-the-road trucking as a serious part of our business,” Bongers says. That happened a year ago when its system was updated to be less costly and less cumbersome inside a truck cab.

He declined to name interested U.S. fleets but says one represents more than 20,000 trucks and that SmartCap’s system also is being tested by fleets ranging from five to 100 trucks. In addition to clients in mining, SmartCap has trucking clients on other continents.

Automotive industry embracing fatigue monitoring tech. Free coffee, anyone?
By Max Heine

Unlike trucking, where fatigue-monitoring vendors are trying to sell their wares to fleets, in the automotive world it’s vehicle manufacturers who are the main drivers in adopting the technology.

A March New York Times article surveyed technologies being used or in development by Mercedes, Volvo, Audi and automotive suppliers. The Times found that in most cases, the systems focus on vehicle movements, such as lane departure, via road-facing cameras and the car’s electronic control unit data. Driver-facing cameras are also in the works.

Cars are a little ahead of Class 8 trucks in autonomous driving technologies, and some manufacturers are weighing the challenges of having the car intervene on its own when a driver gets too tired.

For example, Nvidia, which supplies chips to Mercedes, Tesla and others, is developing a system that will learn a driver’s typical physical attributes when driving, such as head position, blink rate and steering style. When notable departures from those norms occur, the system would warn the driver, possibly even seizing control and safely parking the car.

Volvo uses similar input in its Driver Alert Control that’s standard on its XC90, a luxury SUV. “If it senses that you’re driving more erratically, or there are signs that you’re drowsy or distracted, it’ll give you an audible warning and a message on the instrument panel suggesting you take a break,” Volvo says on its website.

Volvo also has a grant to study how effective it would be to offer the driver a free cup of coffee and directions to the nearest coffee shop when drowsy driving is detected, the Times reports.

Bosch, which supplies many automakers, is developing a fatigue-detection system that tracks things such as head and eye movements. It also would go beyond camera data to integrate biometric indicators such as heart rate and body temperatures, the Times reports.

Biomonitoring of car drivers had plenty of attention in Europe through the HARKEN project. Conducted largely between 2000 and 2011, it involved European research centers and businesses with an interest in measuring a driver’s level of fatigue without using cameras. The technologies focused mostly on measuring heart rate and respiration, using special textiles and sensors in the seat cover and seatbelt.

Such systems could capture the needed data “in a completely unobtrusive manner,” and testing showed its potential for being integrated in cars, concluded a technical paper summarizing the project.

Notable in its absence in the automotive sector is wearable technology. That’s no surprise, since noncommercial drivers, being subject to much less regulation than truckers, would not be inclined to bother with donning gear every time they get behind the wheel.

Source