Statistician Jim Hedlund can see you while you’re driving.
Well, maybe not exactly you, but a pretty good approximation.
The Strategic Highway Research Program (SHRP) II – a study by the Transportation Research Board (TRB) of the National Science Foundation (NSF) Hedlund consults on– has fortified the vehicles of over 3,000 subjects with advanced surveillance tools. These include lane trackers, speed recorders, and video cameras on the front and back of the vehicle and on the driver’s face.
For the past seven years, SHRP II has been watching. The data they collect will be used to extrapolate what you – and the legions of other drivers – are actually doing behind the wheel. For the traffic safety industry, a study this size utilizing this type of technology could be ground breaking.
Until this point, naturalistic driving studies – our main resource for understanding how people drive and how drivers should be regulated — have fallen under two categories: Crash data reports based on police records taken after the accident, and university and government funded simulation studies that rely on virtual reality to approximate behavior. In an age that boasts imagery intelligence and virtual machines, those methods could be considered primitive.
Hedlund — who has a PHD in mathematics from University of Michigan and is the former Chief of the Mathematical Analysis Division at the National Highway Traffic Safety Administration (NHTSA) – is excited. After 37 years of informing policy makers and overseeing enforcement campaigns, Hedlund and his peers may finally have answers. Are people using cellphones when they crash? How do they react when the car in front comes to a sudden halt? If a kid runs into the street, at what point does the driver respond ?How do drivers negotiate sudden storms?
Thus far, these kinds of questions have been addressed speculatively. Hedlund believes that SHRP II – which is slated to wrap up in November and publish findings in December 2014 – will shape traffic safety efforts for the next decade and a half, if not longer.
“This is the next big thing,” says Hedlund, “I think it will be quite useful – perhaps enormously useful – in helping understand driver behavior and helping to make our roads safer.”
Throughout his career, Hedlund has been involved in several ground breaking research projects including the first National Automotive Sampling System (NASS) and the first Fatality Analysis Reporting System (FARS), the bed rock study in traffic trends and fatalities.
He’s presided over countless efforts including Partners in Progress, which strategically reduced drunk driving, and the National Automobile Dealers Association (NADA) Air Bag and Seat Belt Safety Campaigns of the late ‘80s and ‘90s. Since retiring from NHTSA, the Ph.D. Mathematician has published over 80 papers on traffic safety research and compiled the manual, Countermeasures That Work, which has become the bible of highway safety patrol.
Along the way, he’s helped move traffic safety practices from logic to science.
Recently, Hedlund was awarded The James J. Howard Highway Safety Trailblazer Award by The Governor’s Highway Safety Association for his lifetime of leadership.
Hedlund’s first job in traffic safety – the one that pulled the former professor from academia – was statistician for the National Center for Statistics and Analysis (NCSA), then a new division of NHSTA. The goal was to bring a wide range of statistical support and analysis to traffic safety laws and mandates. At the time NCSA was launching FARS and NASS – among the first nationwide statistical analysis on traffic safety. Both operations are updated annually and widely cited. FARS acts as census on traffic accidents. NASS, meanwhile, looks at a representative sample of accident reports to describe safety factors.
“Those looked like really interesting data sets to get and analyze and figure out what was working, how to prevent crashes and how to reduce injuries,” says Hedlund. “It made perfectly good sense from a public health perspective.”
It’s easy to take for granted that high-level decision making starts with data. With traffic safety, however, that’s not always the case. Innocent lives are on the table and policy makers and traffic safety officials can face tremendous pressure to act. It’s easy to approach the process intuitively. If cellphones are problematic, ban them. If more people move onto a street, lower the speed limit. If bikers collide with traffic, corral them.
Think about this: Since 2008, bans on cell-use while driving have been adopted by many states. Yet, The Highway Loss Data Institute recently found drivers are as likely to get hit in states with bans as those without.
Understanding the breach between safety practices and human behavior – and figuring out how to close it – is what Hedlund and others in his
Image source: versageek on Flickr, http://www.flickr.com/photos/versageek
field have been working on. “It’s all very simple to think, Oh we have problem like distracted driving or cellphone use, let’s just pass a law,” says Hedlund. “Does the law make any difference? That’s a statistical, analytical question.”
In the ’80 and ’90s, while he was an associate administrator for behavioral programs at NHTSA, Hedlund helped determine – and popularize – the laws and techniques that do work. In the mid ’80s, New York became the first state to issue a mandatory seat belt law. Other states followed – all now have mandates except New Hampshire (coincidentally the “live free or die” state). The use of seat belts seemed to increase, but not enough.
At the time, Hedlund was administering NADA’s National Air Bag and Seat Belt Safety Campaign. The program helped support Click It or Ticket, a media-hyped campaign that publicized law officers targeting seat belt offenders. To measure the campaign’s effectiveness, data collectors stood along roadways with Belvue screens and counted seat belt wearers.
The campaign seemed to work. By the mid ’90s, seat belt use increased by 50 percent. It’s not just the law, Hedlund proved, but the combination of law and enforcement campaigns that create actual change. For these campaigns to be effective, however, people need to know they’re happening. That means broad coverage and strong messaging.
It can also means tailoring the message to specific audiences based on demographics including geography. People in rural parts of the U.S., for example, tend to have different world views and responses than people in New York City.
“Most ad agencies will tell you, you need to reach different people in different ways,” says Hedlund. “That strategy has been adopted in much of traffic safety as well.”
“Checkpoint Tennessee” followed the “Click It or Ticket” strategy. When the state laid out sobriety checkpoints – and then publicized them – drunk driving dropped. Hedlund led the evaluation that proved the strategy’s effectiveness inspiring other states to follow. Now checkpoints are a regular practice through out the United States.
In 1990, Mothers against Drunk Driving (MADD) launched “20 by 2000” with the goal of reducing alcohol-related fatalities by 20 percent within the decade. Hedlund’s support of MADD helped establish the national 21-plus legal drinking age and the standard 0.08 percent legal limit for Blood Alcohol Content (BAC), recognized by most states. Those laws, combined with MADD’s gripping advertisements – which showed families victimized by drunk drivers – created remarkable change. By 1995, MADD reached their goal. By 2000, drunk-driving fatalities had been reduced by 32 percent.
By nature, vehicle traffic is difficult to regulate. There are over 250 million registered passenger vehicles and nearly 3 million miles of paved road in the United States. Cops can’t be everywhere. Many highways are now equipped with sensors that measure speed and traffic signals that contain cameras. Still, law enforcement struggles to reign-in seat belt, cell and sobriety offenders.
With increasingly limited resources, state highway safety offices have to be shrewd. They must seek to understand which efforts have real socio-cultural impact and are worthy of investment. That’s why the work of Hedlund and his peers has been so critical.
“If you don’t ask those questions and if you don’t get the data to answer those questions, you will keep on doing silly things, and spending money on things that don’t make any difference,” says Hedlund. “Use the data to show what works and what doesn’t work.”
Countermeasures that Work did precisely that with close to 90 recommendations curated from the larger body of traffic safety research. Topics include educating drivers on the effect of prescriptions drug and the proper use of booster seats. The manual is published by NHTSA and gets updated every year.
The last study of SHRP II’s size and scope, The Indiana Tri-Level Study, was published in 1979. It was successful in determining over 90 percent of vehicle accidents are caused by human error. Yet, considering advances in vehicle technology, highway surveillance, changing socio-cultural trends, and the spread of mobile technology, Indiana Tri-Level can tell us little more about why we crash. When SHRP II is published next year, the technology used to understand traffic safety will have finally caught up. We will know, precisely, what those human errors are.
That kind of information, according to Hedlund, is the first step in changing the safety of our roads. “The goal isn’t to catch people and give them tickets,” says Hedlund. “The goal is to change behavior.”
Throughout his career, Hedlund has been involved in several ground breaking research projects including the first National Automotive Sampling System (NASS) and the first Fatality Analysis Reporting System (FARS), the bed rock study in traffic trends and fatalities.
He’s presided over countless efforts including Partners in Progress, which strategically reduced drunk driving, and the National Automobile Dealers Association (NADA) Air Bag and Seat Belt Safety Campaigns of the late ‘80s and ‘90s. Since retiring from NHTSA, the Ph.D. Mathematician has published over 80 papers on traffic safety research and compiled the manual, Countermeasures That Work, which has become the bible of highway safety patrol.
Along the way, he’s helped move traffic safety practices from logic to science.
Recently, Hedlund was awarded The James J. Howard Highway Safety Trailblazer Award by The Governor’s Highway Safety Association for his lifetime of leadership.
Hedlund’s first job in traffic safety – the one that pulled the former professor from academia – was statistician for the National Center for Statistics and Analysis (NCSA), then a new division of NHSTA. The goal was to bring a wide range of statistical support and analysis to traffic safety laws and mandates. At the time NCSA was launching FARS and NASS – among the first nationwide statistical analysis on traffic safety. Both operations are updated annually and widely cited. FARS acts as census on traffic accidents. NASS, meanwhile, looks at a representative sample of accident reports to describe safety factors.
“Those looked like really interesting data sets to get and analyze and figure out what was working, how to prevent crashes and how to reduce injuries,” says Hedlund. “It made perfectly good sense from a public health perspective.”
It’s easy to take for granted that high-level decision making starts with data. With traffic safety, however, that’s not always the case. Innocent lives are on the table and policy makers and traffic safety officials can face tremendous pressure to act. It’s easy to approach the process intuitively. If cellphones are problematic, ban them. If more people move onto a street, lower the speed limit. If bikers collide with traffic, corral them.
Think about this: Since 2008, bans on cell-use while driving have been adopted by many states. Yet, The Highway Loss Data Institute recently found drivers are as likely to get hit in states with bans as those without.
Understanding the breach between safety practices and human behavior – and figuring out how to close it – is what Hedlund and others in his
