Why A Mining Company Is Getting Into Face Recognition Software

Drowsy driving is notoriously tough to detect. There’s no test to prove it, the way a breathalyzer can prove someone was driving drunk. But technology to detect drowsy driving is in the works.

In commercial transport, one industry is leading the way: mining. The stakes are particularly high in this field since the enormous haul trucks used in mining are several times the height of a person. Imagine dozing off at the wheel of one of these . 

Caterpillar Safety Services, a consultancy branch of the global mining company, has partnered with the tech company Seeing Machines to put fatigue detection software in thousands of mining trucks around the world. The software uses a camera, speaker and light system to measure signs of fatigue like eye closure and head position. When a potential “fatigue event” is detected, the system sounds an alarm in the truck and sends a video clip of the driver to a 24-hour “sleep fatigue center” at Caterpillar headquarters in Peoria, Illinois.

At that point, a safety advisor contacts them via radio, notifies their site manager, and sometimes recommends a sleep intervention.

“This system automatically scans for the characteristics of microsleep in a driver,” Sal Angelone, a fatigue consultant at the company, told The Huffington Post, referencing the brief, involuntary pockets of unconsciousness that are highly dangerous to drivers. “But this is verified by a human working at our headquarters in Peoria.”

Caterpillar has a four-year license from Seeing Machines to manufacture the software. For now, it’s the exclusive provider of this technology within the mining industry. Some 5,000 vehicles a combination of Caterpillar’s own trucks and those of other mining companies carry the equipment. There are about 38,000 haul trucks worldwide, by Caterpillar's estimate, so the fatigue-detecting trucks are still a small fraction of that, but Caterpillar hopes to eventually equip all of them.

When a “fatigue event” is recorded, it’s up to the mining site to recommend a course of action to the driver, or vice versa. Last month in Nevada, for instance, a mining truck driver had three fatigue events within four hours; he was contacted onsite and essentially forced to take a nap. Last February in North Carolina, one night shift truck driver who experienced a fatigue event realized it was a sign of an underlying sleep disorder and asked his site management for medical assistance. (Caterpillar has mining operations globally from China to Canada).

“It’s not unusual for someone to lose their frame of reference of what is normal in regard to fatigue,” said Angelone. This may be because miners’ shift work goes against typical human circadian rhythms . A driver’s shift is either eight or twelve hours long, said Angelone, but those shifts can occur during the middle of the night, late afternoon or any other time.

“Many sites run a 24/7 operation,” he said. “These drivers are not always sleeping through the night.”

In the past year, since the company started recording fatigue events last July, it has recorded about 600 instances, said Angelone. He said this constitutes a stunning 80 percent reduction in fatigue events from previous years.

The biggest reason for this, said Angelone, is that once an alarm goes off in a truck, the driver becomes much more aware of their fatigue, and is more cautious and proactive about drowsy driving than they would be otherwise. 

These results invite the question of why fatigue detection software has not yet reached consumer vehicles.

One explanation is that the car industry has not been slow to embrace the technology, but that commercial trucking has been particularly fast.

“There is a lot of incentive to improve safety in our industry,” said Tim Crane, general manager of Caterpillar Safety Services. “Our vehicles are huge and pose unique challenges, so the government really wants to see that we’re trying.”

Crane expects the use of fatigue detection technology in consumer cars to increase “exponentially” in the next few years. Jeremy Terpstra of Seeing Machines echoed the sentiment.

“We have arrangements with many different car manufacturers,” he said. “It’s only a matter of time before this technology is in all vehicles, everywhere.”