Employing AI and Machine Learning to Reduce Repair Costs

Wheelset repairs account for more than 50% of most equipment owners’ maintenance spend each year. Yet, owners have little control over how and when most wheelset replacements happen, from choosing new or turned wheelsets to what brand of components are used.

Current Strategies to Manage Maintenance Costs

Because of the high cost of wheelset repairs, equipment owners have long sought an improved strategy for maintenance planning and management. Owners often employ the following strategies:

• Developing in-house repair facilities and attempting to route cars home when maintenance is needed
• Contracting with a repair shop network and mobile services providers to negotiate better repair rates
• Establishing relationships with component providers in hopes of gaining access to specific components

Major Cost Drivers and Challenges

An underlying issue remains with each of these strategies — without visibility into when a component will need to be replaced, equipment owners have little to no ability to route their cars to a home or specific shop. With current maintenance planning strategies, the following challenges prevail:

• Difficulty accessing equipment to effect repairs
• Difficulty accessing preferred components when repairs are performed by roads
• Inability to avoid foreign wheelset repairs at AAR rates

Because wheelsets account for the majority of maintenance spend, the potential benefits of improved wheelset maintenance control are significant. Working with equipment owners, TransmetriQ has developed predictive intelligence to identify the crucial missing information from all current strategies: when wheelsets will fail.

How it Works: Employing Component Tracking Data with AI and Machine Learning

Trained on data from the Railinc Component Tracking program and utilizing insights from millions of wheelset lifecycles, TransmetriQ’s AI Models predict the probability that a wheelset will reach condemnable status within a mileage window. 

For equipment owners, these predictions will enable identification of wheelsets with no current actionable alerts that are at an elevated risk of reaching condemnable status soon. With this visibility, car owners will have the opportunity to route cars to a preferred shop for repairs and optimize repair decisions when a car is already at the shop. With the knowledge of when a wheel is likely to need replacement, equipment owners can choose to make repairs now instead of risking foreign repairs. This can enable changes to planned maintenance that help improve equipment availability and proactively manage maintenance spending.

Similarly, repair shops can also leverage these predictive analyses to identify repair opportunities. By employing the TransmetriQ service to check contracted equipment, shops can identify at-risk cars and work with car owners to address the issue.

For railroads, improved visibility into remaining wheelset life has the potential to enable greater control over wheel repairs on their owned and leased fleets. For example, if a predictive scan identifies an at-risk car, a road can have the wheel changed at a home shop rather than by an interchange partner at AAR rates.

Ultimately, visibility into remaining useful wheelset life will allow equipment owners to maximize equipment availability, manage the risk of incurring foreign repairs, and optimize their maintenance investment.

Get Started Leveraging Predictive Analytics

TransmetriQ’s team of experts can help jumpstart your team’s journey into advanced business visibility and optimized decision making. Contact us here to connect with an expert or learn more about Wheelset Intelligence.