Transportation Interview
Testing Tomorrow's Transportation Systems
Interview with Christine Johnson
In 15 years, how will a typical drive be different from what it is today?
Dr. Johnson: There will be more warning, more automated responses, which will probably allow closer headways between vehicles, in turn resulting in more throughput. It will also be quite a bit safer. The car will be capable of monitoring where it is and what it's doing with respect to other vehicles as it travels. I could be wrong, but I am not sure that you're going to see automated cars in cities. There may be some automated truck routes.
Dr. Christine Johnson is the Director of the Multimodal Intelligent Transportation Systems (ITS) Joint Program Office at the U.S. Department of Transportation in Washington, D.C. She has been instrumental in shaping Federal ITS program strategies and policies. Dr. Johnson received her PhD in public policy analysis from the University of Illinois
In 1997, the U.S. Department of Transportation conducted a full-scale demonstration of an automated highway along a 7.6 mile segment of Interstate in San Diego, California. If the technology could be demonstrated then, why shouldn't it be available by 2015?
The automated highway is remarkably technically feasible. The real issue is the dynamics of the system as a whole. Let me explain. Let's say you were able to build an automated highway and you could increase throughput and speed by a factor of four compared to a normal highway. So instead of getting commuters from an outlying suburb into town in one hour, you got them there in 15 minutes. Fine. But where do they go when they take the off-ramp into town? Taking this into account, the only way we are going to see automation is as incremental change. This will begin with warnings to drivers and some control by the vehicle in emergency situations.
No automated highways, but plenty of smart safety systems. That's the prognosis?
I think that within 15 years crash avoidance technology will achieve the level of penetration in the U.S. that car air conditioners have achieved today. Most cars at the mid-level price range will have it. Most cars will also have telematicsthat is, navigation, GPS and more. But the driver will still be in control of the vehicle.
"Drivers" wave both hands as a platoon of automated vehicles glides along I-15 in San Diego in a 1997 full-scale demonstration conducted by the U.S. Department of Transportation
Crash avoidance is largely a matter of sensors. Will the sensors that help tomorrow's cars to avoid accidentscameras, radars, tire and weather sensorstransmit information to nodes that feed pertinent data back to other vehicles?
What you're talking about is what we call the situationally aware vehicle. I believe this is a very exciting concept. Most of the individual technologies are doable. If this concept can be put into effect it will be an exciting approach to the interface between the road, the automobile and the driver. For the first time, we would have a loop rather than simply a collection of drivers. Each vehicle would feed information about its surroundings into a database, and each vehicle would draw information from that database that was important for it.
When you add up the effects of all the safety systems that are on the horizon, what savings in terms of lives, damage and time do you see?
About six million motor vehicle crashes occur in the United States each year, causing about 41,000 deaths, 3.2 million injuries and around $150 billion in damage. Our research indicates that if just three of our Intelligent Vehicle Initiative technologies were deployed on all new vehicles, annual crashes could be reduced by about one in six. We also believe we can get between 15 and 30 % improvement in overall road use efficiency through improvements in speed and reductions in stop-and-go traffic.
Four Problem AreasAnd Their Solutions
Rear-end collisions account for one in four crashes, or over 1.5 million accidents a year. The National Highway Traffic Safety Administration estimates that radar and video technologies now being tested would be effective in preventing 49 % of these crashes (759,000 per year).
Road departure accidents account for 19 % of crashes, or over 1.2 million accidents per year. Lane and road-edge tracking systems combined with "drowsy driver" advisory systems could eliminate up to 458,000 accidents.
Intersection collisions account for about 30 % of all accidents, or some 1.8 million collisions per year. No figures are yet available as to how many of these crashes could be avoided through technology.
Lane change and merge accidents account for one in 25 crashes with 90 % of these resulting from lane changes. It is estimated that vehicle-based video and radar systems could apply to 192,000 of the 200,000 crashes in this category each year.
Considering the fact that more and more elderly people will be driving in the future, how can sensor systems help them to remain independent without jeopardizing safety?
This is a key point. Age will drive tomorrow's market. If you look at who buys high-end cars, it's the people who are 50 and 60 years old. They will be looking for safety features like enhanced night vision, intelligent cruise control, blind spot sensing, and rear-end crash avoidance. Sensor systems will make all of these possible.
What technologies do you see as being the most exciting for the long-term future?
First, the situationally aware vehicle, which we've already discussed, is truly interesting. Second, real-time travel information will fundamentally change the transportation picture. In this country we have focused on the construction of roads. But as we get to the point of easily and continuously communicating traffic conditions to drivers, the focus will change from cement and asphalt to information infrastructures. Right now, in our 78 largest U.S. cities, about 22 % of the roadway system uses cameras or sensors to yield information about what's going on. One of our goals is to improve this level in terms of traffic and weather sensing capabilities. Right now, we have to look at this as an infrastructure. In the future, our cars may be the infrastructure.
What's the status of your testing program?
Right now what we have going is the rear-end crash avoidance operational test. This is a radar-based technology. We are just about to ramp up the number of vehicles involved. We are also testing a camera-based system that helps to avoid run-off-the-road crashes. It constantly analyzes road geometry and notifies the driver if he's straying. That's far more complicated than it appears. One of the techniques involves a precursor to a totally situationally aware vehicle. We are now at the point of putting real people in real cars in real traffic and giving these technologies a very robust test.
Are there significant liability concerns?
What you hear from many corners is that liability is a problem. I am not convinced of that. Automobile company lawyers say that if products meet the reliability tests that Detroit demands, then we shouldn't have any more problems with the new safety products than with any other product. At the moment we are still in the phase of bringing the reliability of rear-end crash avoidance and lanekeeping systems up from 98.9 % to 99.99 %.
Interview conducted by Arthur F. Pease