Contemporary Analysis

Data Science

Grant Stanley

Union Pacific Railroad Predictive Analytics

For most, seeing a Union Pacific diesel-electric locomotive painted in the historic Armour Yellow, Signal Red, and Harbor Mist Gray does not bring predictive analytics to mind. However, the largest railroad network in the United States is showing the effectiveness of analyzing non-conventional data forms to increase operational efficiency.

Applications in industries including financial services, insurance, retail, and healthcare are what commonly come to mind when discussing predictive analytics. Thinking outside the box, Union Pacific is currently using bearing acoustic monitors to identify faulty or failing bearings on trains. The acoustic detectors have been used for years, but enough data has now been collected to show the correlation between failing bearings and certain frequency spectrums. This correlation has been applied to real time acoustic readings to help predict bearing failures and ultimately decrease derailments and costly train delays. Download our Case Study on Mechanical Failure and Predictive Analytics

This example shows predictive analytics can be applied to almost any business in almost any industry. The goal being to make more informed decisions (when to replace bearings on railcars) by understanding trends. To understand what information can even be analyzed is the most difficult obstacle to overcome when discovering new possibilities of predictive analytics. As soon as immeasurables are turned into measurables by quantitatively reducing uncertainty, predictive analytics can be applied.

Union Pacific has saved tens of millions of dollars by taking the time to understand and analyze  the sound of bearings. Next time you see a Union Pacific train plowing through the countryside, imagine the competitive advantage your business would possess if you could make more well-informed decisions using predictive analytics. The possibilities are truly endless.

 

Thoughts? Post a comment.

Categories

See all