A couple of weekends back, the New York Times reported on a unique study into driver behaviour:
THE screams and gasps were just what one might expect from an audience watching a horror movie, but the scenes on the screen were slices of everyday life: real-world traffic accidents in progress, seen in footage recorded inside the ill-fated cars as part of a study of driver attention patterns.
Playing on a digital screen, and eliciting alarm from a crowd of experts here for a conference on auto telematics, were some of the 82 crashes -- and almost 10 times as many near misses -- recorded during a yearlong research project by the Virginia Tech Transportation Institute in Blacksburg, Va. The goal of the study was to collect the kind of information that does not usually turn up in accident reports, insurance claims or other types of after-the-fact data gathering. It found that driver inattention was the overwhelming cause of the crashes in the study.
This was the 100-Car Naturalistic Driving Study, the "first large-scale instrumented-vehicle study undertaken with the primary purpose of collecting pre-crash and near-crash naturalistic driving data".
One hundred cars were fitted out for data collection with:
Five channels of digital, compressed video
Front and rear radar sensors
Accelerometers
Machine vision-based lane tracker
GPS
Vehicle speed sensor
The video system recorded "five camera views monitoring the driver's face and driver side of the vehicle, the forward view, the rear view, the passenger side of the vehicle, and an over-the shoulder view for the driver's hands and surrounding areas".
Being a naturalistic study, drivers (mostly) drove their own cars, received no special instructions and indeed behaved as if they weren't being observed:
There is every indication that the drivers rapidly disregarded the presence of the instrumentation, as is indicated by the resulting database containing many extreme cases of driving behavior and performance including: severe fatigue, impairment, judgment error, risk taking, willingness to engage, aggressive driving, and traffic violations (just to name a few).
Data were collected over a 12-13 month period.
Near-misses missed—until now
The study collected useful data on 69 crashes but also new and important data on 761 "near crashes" and 8295 "incidents" (New York Times):
The incidents in the study include 761 near crashes recorded in nearly two million miles of driving. "A near crash is just like a crash except that somebody did something to avoid it," Dr. Hankey said.
Near crashes are not captured in driving statistics now because they are rarely observed or reported to police. They show an important pattern for driver safety and awareness, though.
And they make for some chilling, real-life horror movies:
Among the gleanings was a nighttime video of a 20-something male driver whose eyes close and head nods as he speeds down a highway. The car drifts toward approaching headlights; the audience screams warnings while the split-screen view shows him drowsing. An oncoming car swerves onto the shoulder and narrowly squeaks by; the sleepy young man is startled awake only after a collision would have occurred.
Your attention, please!
Most of the main findings from the study so far are regarding inattention:
Driver Inattention:
- Nearly 80 percent of all crashes and 65 percent of all near-crashes involved driver inattention just prior to (i.e., within 3 seconds) the onset of the conflict. Prior estimates related to have been in the range of 25 percent of all crashes.
Rear-End-Striking Crashes:
Visual inattention was a contributing factor for 93 percent of rear-end-striking crashes.
In 86 percent of rear-end-striking crashes, the headway at the onset of the event was greater than 2.0 s.
Most near crashes involving conflict with a lead vehicle occurred while the lead vehicle was moving, while 100 percent of the crashes (14 total) occurred when the lead vehicle was stopped. This indicates that drivers are sufficiently aware and able to perform evasive maneuvers when closing rates are lower and/or expectancies about traffic are not violated.
And it's comes as no surprise to learn that
The sources of inattention that generally contributed to the highest percentages of events … were wireless devices (primarily cell phones) internal distractions, and passenger-related secondary tasks (primarily conversations).
Cyclists at risk?
Of all the crashes, near-crashes, and incidents pedestrians and cyclists featured in no crashes and relatively few other events: 6 pedestrian near-crashes and 108 pedestrian-related incidents, and only 16 cyclist-related incidents.
But despite those pleasingly low numbers I think there are some worrying factors here for cyclists. That inattention is a causal factor in 93% of the recorded rear-end collisions suggests that cyclists are at particular risk of being inadvertently run-down from behind. Anecdotally, we already know this – and it must send a shiver down the spine of every road cyclist to see people using their phones while driving.
And also driving while drowsy "increases an individual's near-crash or crash risk by four to six times" and "was a contributing factor for 22 to 24 percent of the crashes and near-crashes." Or as the New York Times put it:
Fatigued drivers were even more dangerous to themselves and others, the study found. Roughly 46 percent of accidents
and incidents recorded during the study involved some form of fatigue, with a surprisingly high number occurring during morning commutes.
[emphasis added]
I think the implications here are obvious.
Limitations
The authors of the study recognise the limitations of their work:
the study was small with the presence of 15 police-reported and 82 total crashes, including minor collisions. Furthermore, drivers were represented from one area of the country… One purpose of a large-scale study would be to have a statistically representative sample of crashes (perhaps 2,000) and a more representative driver/environment sample.
So it pays not to generalise too much from this. However, at the moment this must be a pretty unique and rich data set and as the database is going to be made publicly available on the Internet, I'm sure there are many more findings yet to be drawn from the 100-Car Naturalistic Driving Study.