(The following article by Christopher Dinsmore was posted on the Virginian-Pilot website on March 21.)
NORFOLK, Va. — The railroad industry is developing such systems for its locomotives in six places around the country. Known as positive train control, the systems are being designed to improve the safety and efficiency of railroads.
Such computer-based systems could replace the signal light-based system that now controls operations on much of the nation’s railroads.
Those involved say the system promises to revolutionize the industry as switching to diesel locomotives from steam trains did in the mid-20th century.
“What positive train control is, quite frankly, is as much a revolution as going from steam to diesel,” said John M. Samuels, senior vice president of operations, planning and support at Norfolk Southern Corp., the railroad based in Norfolk. “It’s that much of a transition. What it is, in essence, is developing the digital railroad.”
While positive train control promises to make railroads even safer than they are today, it is a complex system that will take time to develop to the point where it could be used commercially.
Even then, the investment required to install the system will be immense and possibly not cost-effective. Railroads will have difficultly justifying the expense.
“The railroads have spent hundreds of millions on this,” said Tom White, a spokesman for the Association of American Railroads. “The hope is that you can come up with a technology that increases safety and increases efficiency and is cost-effective. But no one can find one that’s cost-effective.”
Current estimates suggest that the cost of implementing positive train control just along the main lines of the nation’s railroads to be between $4 billion and $6 billion, Samuels said.
The industry as a whole spent nearly $5.9 billion in 2003 on capital expenditures, including equipment and roadway maintenance and construction.
Samuels is chairman of the program management committee steering the North American Joint Positive Train Control Program, which is testing a system on Amtrak locomotives over 110 miles of Union Pacific Railroad track in Illinois.
Started in 1998, the program is a joint venture of the Federal Railroad Administration, the Illinois Department of Transportation and the Association of American Railroads. Each put up $20 million in funding.
The program is being managed by the Transportation Technology Center Inc., a subsidiary of the railroad association.
There are five other positive train control systems being tested in various locations, but all are still in development.
“There is no positive train control system that is truly a digital system,” Samuels said. “They are overlay systems that work in conjunction with the existing signalling system.”
As these projects have been developing, the Federal Railroad Administration has been writing the regulations to establish minimum performance requirements. The FRA issued the final rule on March 7, the first revision of railroad signalling and train control regulation since 1984.
“The safety benefits of PTC technology are real, and its accelerated implementation holds the promise of reducing certain types of railroad accidents,” said Robert D. Jamison, the FRA’s acting administrator, in announcing the new rule.
The FRA said such systems must improve safety by automatically controlling train movements and speed, including bringing a train to a stop when needed.
Positive train control “can prevent collisions between trains, provide warning of other on-track equipment or hazards and enforce track speed limits. Future capabilities may give motorists in-vehicle advance warning of a train’s approach at highway-rail grade crossings.”
The FRA said any such system must meet or exceed the safety of existing signal and control systems.
“This is really a test bed for these standards,” Samuels said of the Illinois project.
Samuels likened the existing train control system used by railroads to that used to control motor vehicles on the roads. Just as roads have speed limits, so do sections of railroad track. Signal lights regulate train traffic just as traffic lights control intersections.
But unlike vehicle movement on the roads, train movements are managed by a central dispatcher who keeps track of where various trains in a designated territory are operating. The dispatcher switches lights out on the system to regulate the flow of the trains. However, the engineer controls the train’s speed, and must obey speed limits and restrictions as well as the signals.
“We have smart people driving dumb machines, obeying lights,” Samuels said.
Speed control is one of the big issues for railroads. Unlike a car exceeding the speed limit on the highways, a train going too fast will more than likely derail, he said. However, compliance with speed limits is vested in the engineer.
“In positive train control that’s all vested in the system, and the system won’t allow an engineer to speed,” Samuels said.
If the system detects the train exceeding the limit, it warns the engineer. If the engineer doesn’t slow the train, the system will automatically stop the train. Samuels calls it failing safely.
A railroad following its operating rules should be safe, but the current system allows for human error, Samuels said. Positive train control would greatly reduce the risk of human error.
The system relies on knowing where every train is at all times, so each locomotive is outfitted with a Global Positioning System-based tracking system that is backed up by gyroscopes and accelerometers to track where the train is from its last good reading by a GPS satellite.
“All that refinement can increase the capacity of your railroad,” Samuels said.
A problem on one train could be instantly relayed to trains in its vicinity, allowing them all to slow or stop simultaneously. That would allow trains to operate more closely to one another and therefore increase the railroad’s capacity.
Instead of just seeing a speedometer and some other limited information, a locomotive engineer would have a screen that shows a variety of information about the upcoming track. It would show the train’s current speed, the speed limits and restrictions ahead and the status of upcoming switches and perhaps even highway grade crossings. It would show where there were work crews out on the rail line. Even the contour of the land will be displayed, since that can effect the train’s speed.
The locomotive’s onboard computer would be in constant contact with the dispatching system, so that it can be updated instantly.
“What happens here is, if you’re a locomotive engineer, you can see seven miles out,” Samuels said.
And if something goes wrong, the train warns the engineer, and if he or she doesn’t respond, the system stops the train.
Norfolk Southern is closely watching the system for its potential and participating in its development, but the actual commercial use is still years away, Samuels said. The plan for the Illinois system is to start using it in so-called revenue service by the end of the year on Amtrak trains.
Others won’t even consider rolling it out until all the kinks are worked out. “We can’t afford to have a system that’s stopping trains left and right,” Samuels said.
He figures the system needs five to seven more years to incubate, then it could be implemented over the next 10 years by about 2020.
But its cost may prove to be the biggest barrier.
A positive train control system involves not only the software, but also hardware for the locomotives, dispatching centers and other offices, and in the field.
“This is a very expensive system,” Samuels said.
The Illinois system’s software is being developed with an open architecture that will be owned by the railroad association to help reduce the cost and allow it to be adapted to suit specific railroads. It’s also being developed with non proprietary, off-the-shelf equipment.
“We’re a very capital-intensive industry,” Samuels said. “And we, as an industry, are not earning our cost of capital. Our return on investment does not exceed our cost of capital. We wanted to control our costs.”
The Federal Railroad Administration acknowledged in a letter to Congress about the system’s development that positive train control likely wouldn’t be cost-effective for railroads. In fact, the FRA concluded that the increased safety and efficiency gains wouldn’t pay for the cost of positive train control in 20 years, said the railroad association’s White.
“Railroads are so safe that the cost of accidents doesn’t justify the expense,” Samuels said.