How Artificial Intelligence Will Change Freight Rail Maintenance

Railroads play an essential role in today’s global economy.  As the most efficient land-based mode of transport for freight and the most reliable commuting method for passengers, both freight and passenger rail enable economies to operate reliably, safely and cost efficiently. Given the global pervasiveness of the railroads, making this transportation mode even more reliable and efficient is of paramount importance.

Emerging technologies such as Machine Learning and Big Data are promising candidates to unlock greater utilization by changing existing maintenance practices towards condition based maintenance. Both, mechanical assets such a locomotives and railcars as well as infrastructure assets such as tracks will benefit from this change, as maintenance and impending failure become predictable. Existing successful implementations include monitoring track geometry/rail condition, brake systems, and wheel health to name a few.

While the technology is advancing, railroads have remained steadfast in their established maintenance practices. Great emphasis remains on subject matter experts whose decades of experience are used to manually establish static rules and fault limits instead of incorporating them in data driven Machine Learning algorithms. Widespread adoption and the benefits of efficiency improvements through Machine Learning have thus eluded the railroad industry.

In this talk, I will explore some of the underlying causes of the status quo and how stakeholders can justify the capital expenses required for gradually upgrading their operations to incorporate findings of the fourth industrial revolution. I will furthermore discuss the structural changes required to institute a state of the art program and attracting and retaining necessary talent to enable these advances.