Blogs

Predicting operational failures: Better insights detect problems before they cost you time and money

Post Comment
Business and Technology Writer

For transportation leadership, actionable insights empower operations success on every level. Thanks to data analytics, predicting operational failures down to the details of an individual vehicle is now a possibility. This approach to transportation operations is saving companies money while making the industry safer.

Consider these statistics: The Wall Street Journal notes the price of trucks has increased 60 percent since 2008, incentivizing companies to maintain their vehicles for as long as possible. Further, Mass Transit magazine reports the U.S. has a $85.9 billion transit backlog that will take $2.5 billion in annual investments to maintain.

With these financial pressures on the industry, proactively fixing operational problems can have a significant impact on an organization's bottom line. Data analytics allows transportation companies to prepare for and mitigate losses created by broken trucks, trains and planes. In the following examples, see how the applicability of predictive analytics to fleets, railroads and critical flight components is enlightening the experts who use it.

Predicting failure in the trucking sector

A delivery or rental company's vehicles are the lifeblood of the organization. Understanding how, when and under what circumstances an engine might overheat or an axle might break has implications for entire automotive and trucking fleets.

To that end, companies may want to look beyond a confined set of traditional parameters, such as time and miles, to assess when an engine component might break. In-vehicle sensors are empowering transportation and shipping managers to predict problems more accurately. Jim Beach, technology editor at Trucking Info, points out savvy businesses are coupling equipment-related, real-time information with driver-generated data such as digital messaging about navigation, rerouting, deliveries and even fuel card use. With better insights, operations executives can compensate for maintenance by having backup vehicles ready and plans to redeploy drivers to cover unmanned routes. Data analytics can even allow companies to plan optimized routes and prevent failures in the first place.

Railways leverage data to avoid disruptions

When transportation relies on a complex and interdependent infrastructure, as is the case with trains, the ability to predict operational failures is crucial. Downtime affects entire systems, causing costly rerouting and expensive allocation of additional resources.

In the United Kingdom, where, as Engineering and Technology recently pointed out, long runs of harsh weather can wreak havoc with signaling systems, train cars and tracks, experts are pursuing a policy of predictive maintenance. By collecting massive amounts of data from sensor-equipped train cars, rail companies can not only forecast potential breakdowns, but they can also institute overhaul, replace or upgrade schedules. Data analytics tell engineers the optimal way to plan for maintenance, thereby avoiding the disruptions a mechanical failure could cause.

Airlines tap sensors to make planes safer

One way to assess the breaking points of materials used in complex vehicles is to subject them to a great deal of vibration, repeated weight-bearing experiences and other similar tests. These factors are endemic to aircraft, and in the maintenance repair and overhaul (MRO) space, airlines have been engaged with part-failure analysis for some time. But MRO engineers and operations executives want more than pass-or-fail recommendations from the data they analyze.

As Aviation Week reports, airlines are pushing for even more granular assessments. For example, a valve opening even a tiny bit too fast or slow can inform critical decision-making about replacing or adjusting parts. MRO professionals must be able to quickly analyze and interpret this data and implement changes because, as the Federal Aviation Administration notes, a one-hour delay costs an airline between $1,400 and $4,500.

It's the kind of ahead-of-time approach researchers in U.S. Army labs are applying to aircraft, going beyond the overall health of the vehicle's systems and digging into microscopic changes to metals and materials over time. Data scientists and mechanical engineers are increasingly able to detect nano-sized wear and project precursor conditions that aircraft maintenance staff can look for in terms of part fatigue and failure at the material level.

The underlying strategy in these situations is one of empowerment: Greater insight means greater foresight, and greater foresight allows transportation companies to implement strategies for predicting operational failures that disrupt infrastructure. Data analytics are fueling this transformative approach, helping the industry to maintain (and eventually shrink) its huge infrastructure backlog.

Visit IBM's Travel and Transportation Industry Solutions page to unlock the business potential of your travel data.