The 5 Challenges to Adopting a Lean Repair Shop

As we saw in Part I, leveraging lean principles in the rail shop brings greater reliability across the rail ecosystem. By making failures on railcars a predictable, addressable, and integrated part of operating activity, railways can drive significant cost savings and optimize on-track performance. McKinsey estimates that integrating condition-based and predictive maintenance throughout the rail sector would reduce maintenance costs by 15-25 percent.

With rolling stock OEMs and rail operators adjusting to depressed freight demand, getting more out of the same fleet represents an ever-important source of differentiation. Across the industry, consistent fuel costs, rolling stock lifecycles, and infrastructure durability leave 40 percent of the total cost structure variable for each operator.

Given the clarity around the value of lean operating models, why haven’t rail repair shops adopted the model? In our deployment with a Class I Railway, a common challenge we heard from reliability leads is that prioritizing service events doesn’t just involve the shop. The entire railway value chain must undergo digital transformation. It can be a daunting task, but the linchpin of a digitally-transformed railway is a rail shop that takes the initiative and confronts these challenges head-on.

Here are five challenges rail maintenance is facing that it must overcome to adopt lean operating principles:

1. Unplanned Service and the Reactive Maintenance Cycle

Repair shops have long depended on a combination of scheduled maintenance and reactive, unplanned maintenance. To get to a leaner shop model that foresees future operating activity as a locomotive experiences power assembly degradation, for example, the shop must move away from preventive routines to procedures that provide an optimized strategy for servicing power assembly degradation. Initiating that process improvement without adding more overhead is even more challenging when unplanned service events don’t allow for maintenance process refinement.

2. Imprecise Alerts from Sensor Data

After fault-codes fire, work orders saddle repair shops with diagnostics, troubleshooting, or repairs. The problem with these alerts is that they are often imprecise and nonspecific. Even when sensor data indicate a specific component-level failure, those data sit in different collection systems and makes condition-based monitoring unfeasible. As a result, many operators cannot provide sufficient lead time to enable shops to anticipate future operating contexts. In one recent engagement, Uptake found that OEM sensor data mirrored actual component-level conditions only 3 percent of the time.

3. Unstandardized Data Management Practices

Making the rail shop lean will require standardizing varied data collection practices, including supervisory control and data acquisition (SCADA), inspection documentation, problem reporting, dispatching, and work order management systems. These collection and cleansing practices complicate how operators structure data internally to deliver effective analytics. Rail executives admit as much: according to a 2016 survey, 75 percent of rail executives believe that IT and OT systems are poorly connected and that this lacking convergence impedes performance management.

4. Conflicting Claims of Data Ownership

In addition, as rolling stock OEMs attempt to provide analytics to guarantee the reliability of their rolling stock, they are finding that access to historic, peer, and design model data within an operating context are necessary — and that they are lacking it. They are also finding, however, that rail operators are not as willing to part with proprietary information which often represents a source of competitive advantage. Which party owns the data, whether OEM or operator, is set to determine how incentives in rolling stock performance inform the development and precision of railcar analytics and reliability.

5. Tying Maintenance Decisions to Financial Outcomes across the Railway

Relevant and precise insights that are easy to take action on are only so powerful as they impact the bottom-line. Right now, many rail operators have a limited view of how their maintenance strategies are impacting financial performance. A combined predictive and prescriptive view of maintenance can lessen the impact of unforeseen repairs on regular maintenance and operations, cutting dwell time, increasing mission readiness, preventing railcar failures and network delays, bundling service events on individual trains, and tying maintenance strategies to a cost-effective regime of service that optimizes service across an entire fleet.

A systematic, diagnostic view of variability — component-level failures on railcars — provides enough lead time for repair shops to adjust parts inventory, staffing, and planning for other service events. A lean rail shop can make maintenance more reliable by eliminating scheduled routines that do not enhance performance out on the track.

For most maintenance leads, performance reliability, availability, and shop efficiency are three key considerations in making repair leaner. Until shops have asset performance management that cleanses data within native IT and OT to deliver advanced analytics, many will remain stuck in a cycle of reactive maintenance with imprecise alerts on rolling stock conditions not tied to operational outcomes. And as long as that is the case, repair shops are trading reliability, and therefore revenue, for inefficient maintenance spend.

Our rail customers have rescued 75 percent of unused or underused data for analytics.