Loveland Company Helps Keep Trains On Track
LOVELAND, Colo. (AP) - A small Loveland company with roots in the railroad industry is making sure trains stay on the tracks around the world.
DataTraks, a five-employee engineering shop off an alley in downtown Loveland, crunches numbers, develops software and creates systems focused on the safety of railroad equipment.
Much of the company's high-tech work focuses on preventing derailments - catastrophes that result in the destruction of rail cars and their contents, railroad tracks and sometimes human life.
"You can pay for a lot of technology with one derailment," said James Bilodeau, president and owner of DataTraks, who founded the company in 2005.
Bilodeau formerly worked at Transportation Technology Center Inc. in Pueblo, a subsidiary of the Association of American Railroads that conducts rail industry research and testing and is DataTraks' largest client.
DataTraks develops software for and helps maintain a system developed by the Transportation Technology Center that listens to passing trains to detect wheel bearings that are in danger of failing.
The system analyzes the sounds that each wheel makes as a train passes a detector - which is equipped with six microphones on each side of the track.
The United States has 16 detectors, including one south of Colorado Springs; Canada and Australia each has one; South Africa has two; and China has 75, he said.
Bilodeau said a typical coal train can have an average of 504 axles, and the Trackside Acoustic Detection System can quickly plow through many gigabytes of data to isolate a particular axle and alert a railroad company that a problem is brewing.
If a failing bearing avoids detection, it will seize up, and the wheel will stop turning, Bilodeau said. "It will melt that axle off the train and derail it," he said.
"It looks like somebody twisted it off like a Tootsie Roll."
The cost of a derailment is enormous, Bilodeau said. In addition to the loss of rail cars and miles of track, a stretch of rail can be closed for days, disrupting traffic and costing businesses money.
Richard Morgan, an engineering manager at TTCI, said in an email interview that the center works with other software support companies in addition to DataTraks.
While such software work constitutes a majority of DataTraks' business, the company is working on another rail safety system that could give it a major boost.
Bilodeau explained that most railroads now use continuously welded tracks, which are welded together on-site rather than just bolted down end to end. The continuous track makes high-speed travel possible and reduces maintenance costs, but it comes with drawbacks as well, he said.
When the sun heats such a track, the metal expands, and with such long stretches of rail, there's nowhere for that expansion to go, and sometimes the track buckles sideways.
The resulting "sun kink" occurs quickly and dramatically, Bilodeau said, and can derail a train if undetected.
Railroads can tell how much stress a track is under by measuring its temperature, he said. If it knows a stretch of rail is becoming dangerously stressed, it sends a crew to cut out the tracks and weld in new pieces.
DataTraks has developed a system that monitors long stretches of track for stress using wireless networks.
In the company's headquarters, engineers have built "the DataTraks Railroad," a 38-foot-long miniature rail line that represents a 1-kilometer-long stretch of real track.
Using heat lamps, the company replicates the effects of the sun on a rail line and can create a sun kink, he said.
Bilodeau said the company has been testing its SafeTraks sensor system for several years in China, which depends heavily on railroads. If the Chinese government were to purchase DataTraks' technology, it would mean a significant expansion for the company.
"We're still crossing our fingers for a big order from China," he said.
- By CRAIG YOUNG, Loveland Reporter-Herald
(© Copyright 2012 The Associated Press. All Rights Reserved. This material may not be published, broadcast, rewritten or redistributed.)
