To improve freight rail safety, Tony Sultana, principal investigator at the Transportation Technology Center, Inc. (TTCI) in Pueblo, Colorado, analyzes massive amounts of freight rail data to help prevent an accident before it can occur.
“There are approximately 12 million railroad wheels in motion around North America in any given day,” he says. “About 400 of them break each year, and sometimes they can cause a problem. We want to find them before that happens. The question is, ‘How?’ ”
Finding that 0.0033%, he says, takes Big Data and what are known as “composite rules.”
A “composite rule” is the rail term for safety protocols that are developed by identifying a combination of factors that can determine if a piece of equipment is near risk of failure. One composite rule might address worn-out wheels, while another might govern faulty brakes, bearings or trucks, or even the safety of an entire fleet.
Composite rules are possible because today’s rail data warehouses store hundreds of trillions of bytes of information about rail equipment — the equivalent of more than a hundred million digital photos. The rules are important because to make freight rail safer, railroads must predict the rare conditions under which equipment or track is most likely to fail. Big Data enables railroads to identify combinations of factors that lead to problems.
Individual factors by themselves might not be predictors of defects, but in combination they could be, Sultana explains. The industry’s use of wayside detectors to analyze rolling stock has created the opportunity to collect and analyze data to find critical combinations of factors.
Individual railroads and AAR started targeting what are known as combination defects. These are the factors that may ultimately make up a composite rule. For example, the wheels of a train can break. Some minor wear and tear is allowable, but the trick is preventing a small problem from turning into a bigger one. That is where Big Data analysis comes into play. By looking for combination defects, inspection engineers are able to pinpoint the factors causing wear and tear problems.
After extensive study, AAR recently issued a composite rule that sets the industry-wide standards for when wheels must be removed before they break.
As railroads gather more data, the ability to proactively ensure safety by developing composite rules will only increase. Scientists will be able to identify other causal relationships to support the creation of new composite rules, which reduce equipment failure. This allows a very safe industry to operate even more safely and maximize efficiency by spotting and addressing issues before they become problems.