The father of two teenage daughters has come up with a mathematical method that would allow a cell phone to recognize texting while driving.
Mike Watkins, who also is the manager of the applied physics group at Pacific Northwest National Laboratory in Richland, worked with his team to come up with a formula that would allow cellphones to recognize the characteristics of a texter who is driving.
That could be useful to parents wanting to keep tabs on their teen drivers or for companies who want to make sure that employees are not texting on company phones and driving on the job.
"How far this goes and where it ends up are not clear," Watkins said. "But I think we've solved the problem of detecting when people are texting while driving."
His group, which specializes in measuring things that are difficult to see, analyzed 10,000 keystrokes of six texters in a normal setting and found the keystrokes to be predictable.
But then they put them behind the wheel of a driving simulator, hypothesizing that their texting would be different enough that they could tell the difference between just texting and texting and driving.
People who are texting "time-share" their attention between driving and texting, Watkins said. They look down briefly and then up again to check the road.
When Watkins compared the keystrokes from driving and nondriving texters, the differences were consistent and quantifiable, he said. Drivers text and pause and text again, without the rhythm of usual texting.
He started the research after seeing a news report with Transportation Secretary Ray LaHood talking about distracted driving. As the father of teen daughters, Watkins already was aware of the popularity of texting.
LaHood has called distracted driving "a deadly epidemic." Text messaging creates a crash risk 23 times worse than driving while not distracted, according to the federal Department of Transportation.
A driver who takes his eyes off the road for 4.6 seconds to send a text while driving 55 mph would drive the equivalent of the length of a football field blind, according to information compiled by the Department of Transportation.
There also have been some high-profile accidents in recent years. In 2008, 25 people were killed in a California train crash caused by an engineer who was texting and operating the train.
Two years later when a barge ran into a tour boat on the Delaware River and killed two people, the accident was blamed on distraction due to mobile devices, including texting.
Watkins has been working on the issue for about three years, using about $50,000 in seed money that came from a Department of Energy fund that uses money from successfully commercialized technology to develop promising new technology.
After he and his team measured the randomness in ordinary texting and compared it to the increased randomness of texting while driving, they also used a robotic texting finger developed by PNNL engineer Ivan Amaya to make sure their measurements were accurate.
The algorithm developed by the team could be converted to a cellphone app with no additional hardware required, Watkins said.
He has applied for a patent and next will take the system to industry to find commercial interest.
Global positioning systems can be used to show that a cellphone was in motion when the holder was texting, but that is all they can show.
"This approach distinguishes the driver from the passenger," Watkins said.
Possible uses might include parents requesting a text message when their teen's phone detects texting while driving.
Companies might want to check cellphones they issue to employees for improper texting, which could be done without seeing the actual messages being sent.
Insurance companies might be willing to give premium deductions to drivers who can use an app to show that they have not been texting while driving, Watkins said.
Courts also might require the app for convicted texters, just as they now require ignition interlocks for convicted drunk drivers.
Or the use could be as simple as providing a cellphone voice reminder to a driver who wants to break the texting and driving habit.
Because the algorithm detects distraction, it also has other potential uses, Watkins said.
"The idea of assessing someone's cognitive state while they're interacting with a machine or other device has many interesting applications," he said. "You could potentially use it to identify fatigue in heavy-equipment operators or to detect an anomaly in grandma's cognitive function due to a stroke."
Other members of the research team include Mike Hughes, Paul Keller and Ed Beck.
-- Annette Cary: 582-1533; firstname.lastname@example.org