Big Data, Optimized Processes, Analytics

Written by GregPhillips, Product Manager, Analytics and Reporting; and Chuck Hieronymi, Director, Commercial Group, Railinc
image description

RAILWAY AGE, FEBRUARY 2020 ISSUE: Effective use of new tools can improve railcar management, utilization and profitability. Rail equipment owners and lessors—key users of railcar data—manage their fleets in a dynamic environment that fluctuates with changing market demands, railroad operating priorities and real-world conditions. Relying on knowledge, experience and diligence, industry fleet managers keep things moving safely, reliably and efficiently.

Still, with the rise of Big Data, those responsible for railcar fleets know they must make effective use of the new tools for better management of their always-changing and sometimes-chaotic business. The question is, how?

Every minute of every day, 1.65 million (according to Railinc’s 2019 North American Freight Railcar Review), revenue-earning North American freight cars are moving, being processed through terminals or undergoing maintenance. Each of those cars is generating data from a variety of components including wheelsets, brake systems, slack adjusters, valves and more. The result is billions of bits of data moving between shippers, railroads, repair shops and other supply chain stakeholders.

Facing such a monumental flood of information, fleet managers must assess the growing array of data management opportunities and develop analytics capabilities that will improve the management of their valuable fleet assets.

Meeting this challenge is critical if fleet operations are to realize potential benefits that include:

• Safer operations.

• Enhanced reliability.

• Higher yields.

• Improved railcar management.

Harnessing the flood of component-level health data, location information and predictive failure reporting pays off with fewer equipment failures—even as the frequency of inspections and shopping are reduced.

What is the first step?

Understand What’s Driving The Data Revolution

To lay the groundwork for future progress, it’s important to recognize what’s behind this data revolution. Four aspects have a significant impact on railcar owners and lessors:

1. More and better data is being generated. Serialization of individual components by the industry’s Asset Health Strategic Initiative is providing a deeper level of data that is being used to reduce mechanical service interruptions, improve inspection quality, and increase yard and shop efficiency.

2. Process improvements are being implemented for railcar data management. Network participants are generally making strides in gathering, normalizing, validating and mining the growing data volume.

3. New tools are leading to innovations in how data is being used. At the top level is utilization of artificial intelligence (AI) and machine learning for more accurate predictive data, enabling machines to identify exception conditions and in some cases take over certain decisions. Railcar managers are also using less sophisticated tools to create tests where utilization hypotheses can be analyzed. For example, the component serialization capabilities described above allow equipment owners to define, control, and test fleets that are configured with specific components, and then observe and analyze the performance of those components in real-world use. Similarly, equipment owners can develop comparative analyses of the performance of specific manufacturers’ components on their equipment.

4. Helping to economically enable all of this are enhanced IT capabilities that are continuously providing improved hardware, bandwidth, throughput and demand surge capacity.

Fleet owners and lessors can benefit by applying their industry expertise to the task of strategically managing all aspects of these new data insights.

Leverage Data Strategically

Internet of Everything (IoE) technology has been, and continues to be, incorporated in many parts of the rail ecosystem. Intelligent and passive devices on rolling stock and at wayside locations have added new streams of data and information. This has laid the foundation for a new generation of advanced algorithms and more capable data analysis software that will make it possible for railroads and equipment owners to further enhance safety, reliability and service to customers.

Railinc, for example, brings true aggregated rail network visibility from 570-plus railroads across North America to rail equipment managers and other industry stakeholders. In turn, this data can be used for enhanced operational and strategic decision-making across a number of software tools.

Every supply chain participant now has the potential to gain access to near/real-time shipment information, as well as to utilize automated intelligence and machine learning for more accurate predictive data. This turbocharged stream of information empowers managers who work inside silos to make better-informed decisions and adapt to changing circumstances. In some cases, machines are making those decisions and will be doing more so in the future.

Because car fleets come in all types and sizes—from tens of thousands in some cases to scores of cars in others—there isn’t a one-size-fits-all data management solution. Holders of large fleets tend to have greater in-house data management capabilities, and many are actively investing in processes, technology and analytics designed for the particular needs of
their operations.

Mid-size and smaller fleet owners and lessors often cannot justify the development of such systems. Nevertheless, the need for action is pressing.

Focus Data on Tactical Management

The strategic benefits of leveraging big data, optimized processes and analytics to improve railcar management rapidly cascades to the tactical level. As a result, there are many ways users can realize tangible and relatively quick results, including:

• Maintenance and repair.

• Fleet size optimization.

• Car storage.

• Utilization management.

• Lease management.

• Safety.

Maintenance of cars for component-related issues has traditionally been managed incrementally, moving through increasing alert levels starting with a Level 1 “Window Open” that advises degradation has started, and moving up to Level 4 for mandatory and immediate action. These levels provide useful information to plan for removing cars from service for maintenance or repair. At the earlier stages, they leave managers to decide if action should be deferred or made immediately.

In the case of one shipper car owner with whom we have worked, annual railcar maintenance expenditures total some $30 million, nearly half of which is spent on wheelsets—typically, the area where most maintenance dollars go. Comprised of wheels, bearings and axles, these component groups have a high value and are subject to significant wear in normal operations. Since a reduction of just 1% in annual maintenance costs translates to $300,000 for this owner, it makes sense to invest in better utilization of the data now being received from wheelsets from the industry’s Component Tracking program.

Such an owner could now benefit from the availability of historical AHSI data to develop predictive analytics on wheelsets and their components. Work is under way to use historical data about the wheelsets on a specific car and compare it with data about the behavior of all identical or nearly identical wheelsets.

This will allow predictions within a greater degree of confidence that a particular component can be operated for a range of additional miles. Managers will be able to develop methods for scoring each component and prioritizing individual cars for servicing.

Fleet maintenance and repair will be scheduled more precisely to improve asset utilization, customer service safety and reliability, while also better supporting business requirements.

More Data, Fewer Repair Shop Trips

Moreover, data that triggers a decision to move a car to shop for the replacement of a wheel can also be accompanied by a review of the data arriving from different sources about the other components in that car. This presents an opportunity to assess the need for additional repairs, preventing another shopping of the car at a later date. For example, such a review could reveal a low-level issue with the brake system, not yet at the stage requiring immediate service. Nevertheless, it might be worth resolving both issues at one time.

More accurate and more up-to-date repair data will also reduce the number of times cars are unnecessarily sent for service because a previous repair was improperly recorded and/or reported. This level of accuracy at the granular level will facilitate better strategic decisions, as well.

Fleet-level analysis will better identify chronic issues that are sometimes missed with more traditional manual reporting and analysis. Once such an issue is known, management can determine if the root problem is with the actual equipment, with a specific shop network, or even with the type of service the cars experience. Fleet data analysis will also enable a more detailed understanding for the performance of specific car types within fleets, of specific age groups, or other parameters, some of which might be unique to just one particular owner or lessor’s fleet.

As more data is available, it will become an essential support for imagining and realizing new ways of doing business, new business relationships and new tools to plan fleet acquisitions, sizing and replacement.

Relationships and scheduling with contract shop networks could become nimbler when analytics and machine learning are deployed, allowing automated intelligence to direct the movement of cars for shopping in the most efficient way.

Boost Productivity, Safety With Better Data

More detailed information about the expected service performance and life of components can allow owners and lessors to handle the same or heightened demand with a smaller fleet, because it will be possible to accurately predict equipment productivity. Not only would this conserve capital, it could also reduce car storage if fleets are more correctly sized.

Finally, and most important, are the opportunities for continued improvements in safety. Machine learning and AI supported by today’s robust data streams may better identify unusual patterns of failure. This information could then be seamlessly provided to industry committees and the experts who are responsible for monitoring specific components, allowing them to more accurately and quickly decide how critical a particular situation is and how to respond.

For rail equipment owners and lessors, the rise of big data and analytics has opened new opportunities for more efficient, reliable and safe operations. Realizing those opportunities is taking focused effort and time, and the benefits are significant. Investing now will make the industry safer, more competitive and more profitable.

About Railinc

Railinc is the railroad industry’s innovative and reliable resource for rail data, IT and information services. The company deploys applications and data that help railroads, rail equipment owners and other industry participants manage their businesses more effectively and efficiently. Railinc is the largest single source of real-time, accurate interline rail data for the North American railroad system.

Railinc’s Advanced Analytics and Car Repair Management Solutions (CRMS) are applications, data and services that can be configured to specific needs and software environments. These platforms and services can help enable higher revenues, lower costs and more efficient operations. Railinc’s RailSight is a suite of applications designed to deliver rail shipment and equipment management data through a flexible framework that can be adapted to support changing business needs. Its services are provided via RailSight Track and Trace, or the easy-to-use hosted solutions RailSight Monitor and RailSight Demand Trace. RailSight is used by leading shippers, equipment owners, 3PLs and transportation management software providers.

Categories: Analytics, Class I, Finance/Leasing, Freight, Freight Cars, Mechanical, News, Short Lines & Regionals, Switching & Terminal Tags: ,