Technology vs. Fatigue
The SafeTrac alerts the driver when his truck moves from the lane center.
Far too many truck drivers die when fatigue overwhelms them while they are behind the wheel. Now technology may be on the verge of saving some of those lives.
Federal statistics indicate that more than 200 truck drivers a year die from fatigue-related accidents. Many more are injured or cause injury and death when they are overtaken by fatigue. To stem this tide, a small army of sleep researchers and scientists work constantly to develop technology that warns drivers of their level of fatigue and of the performance degradation it causes.
According to Dr. David Dinges, a leading sleep reseacher at the University of Pennsylvania, “efforts to develop fatigue management technologies that are useful in the real world are not ready for prime time.” Even though research has been ongoing for years, Dinges believes the science is still in its infancy and “there is no single device as yet,” which has met the scientific criteria for effectiveness and accuracy.
Nevertheless, the field is crowded with dozens of devices, only some of which have application for driving. These devices have provided researchers with three promising technologies on which to concentrate testing. Dinges is strong in his opinion that while these devices may hold the key to future success, help may also come from some as-yet-unheralded device. But right now, the three devices under scrutiny in a joint U.S.-Canadian study involving one American and one Canadian fleet, are the:
SafeTrac is a vision-based lane tracking system developed under DOT sponsorship by AssistWare Technology and Carnegie Mellon University. The unit is a small oblong box that can be mounted on the dash or headliner. It has two readouts that alert the driver when his vehicle moves from the lane center. Dean Pomerleau at AssistWare says there are two distinct functions tied to these readouts. The drowsy driver function tracks a driver’s lane centering in rolling eight- to 10-minute periods and reads out a numeric score of performance on a constant basis. If this “alertness score” drops below a certain number, the driver is audibly warned. Using the same concept of lane tracking, another readout will tell the driver instantly if he approaches the berm or center line too closely.
Pomerleau says the difference between these two functions is time. The drowsy driver alert creates its score by recording a baseline of driver performance and then tracking lane centering against it.
Ed Cleek, a driver for McKenzie Tank Lines, has been using the SafeTrac lane tracker for almost a year. Cleek says he has incorporated the unit into his normal driving routine. “It keeps you on your toes even in situations where your attention wanders and you’re not necessarily tired,” Cleek says. The lane departure function works whenever the vehicle begins to exit its lane, regardless of a driver’s previous performance. There is a false alarm protection readout as well. It keeps alarms from sounding when a driver uses his turn signals and provides visual confirmation that turn signals are in use.
The system uses a camera to detect a variety of signals from the road that delineate lanes. It primarily uses center lane markings, but it is capable of using oil drops, occasional small retro reflectors or snow ruts when lane markings disappear. “SafeTrac operates effectively in over 97 percent of highway conditions and gives false alarms less than once in every eight hours of driving,” AssistWare says.
Cleek says the SafeTrac is an excellent safety tool. “The lane tracker is sensitive to snow and ice and bugs on the windshield,” he says. “It can give false alarms in construction zones where old paint is still on the road. But I’ve gotten so used to it that I miss it when it’s turned off.”
While devices like SafeTrac determine a driver’s performance by monitoring the vehicle, the Perclos camera, developed by Richard Grace at Carnegie Mellon University in Pittsburgh, monitors the driver directly. Perclos, or ‘percentage eye closure camera’, is a dash-mounted unit with a small camera the driver adjusts to focus on his eyes. Unlike the lane tracker, it is extremely portable and plugs into a cigarette lighter.
Grace, now chief executive officer of Attention Technologies, says he looked at many psychomotor behaviors before deciding to focus on eyelid closure. He explored heart rate, brain waves and galvanic skin response as fatigue indicators. “We found that these behaviors were affected by many things besides fatigue and that devices measuring them gave a lot of false alarms,” he says. “The eyelid closure camera works for everyone without false alarms from physiological causes other than fatigue.”
The Perclos eyelid closure monitor’s display shows the driver how far he has traveled with eyes closed.
Another advantage of Perclos, according to Attention Technologies literature, is that it senses a fatigued driver’s level of alertness very quickly and warns him with an audible signal and a readout of the distance he has traveled with eyes closed. Research has shown that drowsiness increases slowly over an hour or more. Perclos provides early warnings during this progression of drowsiness based on the frequency and length of eye closures. According to the literature, “Drowsy drivers often drive with their eyes closed for several seconds without realizing their eyes are actually closed. Eye closures of three to four seconds are common for drowsy drivers.” A vehicle can travel 360 feet in four seconds at 60 miles per hour.
The Perclos camera has four controls adjustable by the driver: volume, brightness, sound selection and sensitivity selection. There are six sounds ranging from a gentle beep to an aggressive klaxon. The visual display shows the driver how far he has traveled with eyes closed. “Seventy to 80 percent eyelid closure is considered closed, meaning that the pupils are covered by the eyelids,” Grace says. “A driver whose eyes are closed to that extent cannot see.”
During two nights of testing with owner-operator Russ Schaefer, the Perclos camera proved its mettle, warning Schaefer numerous times of eyelid closure. The device is loud enough to help maintain wakefulness, and the readout of distance traveled can frighten a driver who believes he has seen the road constantly. Another readout displays the number of eyelid closures and the frequency during the past hour. “One eyelid closure event should convince the driver to get some sleep,” Grace says.
Among technologies currently being tested, only the Actigraph sleep activity monitor, or ‘sleep watch’ provides both a sleep profile and a performance profile. The watch provides a time display and is worn on the wrist. It measures the activity of the wearer, both when sleeping and awake. When a wearer is asleep, the watch relates the measure of activity to a mathematical algorithm based upon a pre-existing standard of sleep. At the touch of a button the watch will then read back to its wearer the length and quality of his sleep as a percentage of available energy. “There is a direct relationship between the amount of movement and the quality of sleep,” says Tom Kazlausky, vice president of ambulatory monitoring at Precision Control Design, the company that makes the watch. This relationship makes it possible for the watch to determine sleep quality. Much like a fuel gauge, the watch will show sleep as a reservoir, which is drained by activity, making the readout easy to understand.
The watch also measures performance and fitness for duty. A sleep reservoir below 65 percent full is considered to be a sign the driver needs sleep and should not drive. According to Precision Control Design, the device “provides a chronological history of sleep periods back 24 hours and estimates ability to perform selected tasks. Sleep and performance are displayed. An arc shows hours when significant sleep has occurred 24 hours back from the current clock time. When the performance screen is activated, the arc predicts the wearer’s ability to perform a specific task measured from 100 units down to 53.”
The watch has a sensor that cancels out the mechanical movement of the truck, for example, and can discern if the watch has been taken off and put in a stable location. According to Kazlausky, “The watch can only be fooled by taking it off and allowing someone else to use it.”
The fuel gauge-style readout of performance level is constantly available by glancing at the watch face, and you can see a percentage display at the touch of a button.
Kazlausky emphasizes the watch is a personal feedback device. “There are no hard numbers to indicate when a person needs sleep,” he says. “As a driver learns how the watch feeds back information, he will be able to recognize what the watch is telling him. Then he can make decisions about his sleep and his driving time.”
In large part the success of fatigue management technology will depend upon driver acceptance. If drivers find the machines useful, such devices could help save lives. The driver must also be willing to use the information he receives from them. A strong education in the recognition of fatigue and its causes is an absolute necessity. Even more necessary is the driver’s willingness to use what he knows to get the sleep he needs.