Across the country, over 600 short line railroads provide crucial first-mile, last-mile connections and manage switching and terminal operations supporting America’s 140,000-mile freight rail network. These railroads vary in size and scale from regional operations like the Florida East Coast Railway, connecting Miami to Jacksonville, to the switching services that facilitate traffic in the nation’s busiest rail terminal area provided by the Chicago Belt Railway, to the much smaller Swan Ranch Railroad serving customers in an industrial park outside of Cheyenne. Short lines vary widely in their size and resources, but all consider what opportunities and challenges come from utilizing AI and are looking for AI tools to streamline and support work both within their organization and from outside vendors. 

Most short lines are lean operations which rely heavily on outside vendors to provide the information and operational technologies required to run a modern railroad. Additionally, short lines have a lighter traffic density, compared to the six larger Class I railroads. Both these factors make them valuable partners for real-world testing of AI-based technologies.  

One startup, Intramotev, is focused on developing a system that will modify standard railcars, typically used to move bulk commodities, to allow for their autonomous movement utilizing battery-electric, self-propelled axles. Parallel Systems, another startup, is working on a self-powered chassis solution to move standard intermodal freight containers.  

While these two companies are designing unique products to serve different rail freight markets in a similar manner, both are taking advantage of AI to help safely govern the autonomous movement of freight and are testing their technologies on short lines. Once AI technologies finish development and reach the market, short lines are likely to be among the first railroads to take advantage, given the potential for the autonomous movement of individual or small multi-car groups to support the traditional short line business model of serving small-scale customers on an as-needed basis. Automation will enable short lines to more efficiently aggregate their shipments for interchange into the high-density, longer-distance Class I railroad network. 

Future potential for automated operations is not the only way AI is impacting how short lines run their railroads. Wi-Tronix, whose focus on AI includes applications to automate grade crossing equipment inspection and detect the use of mobile devices in the locomotive cab, is among many companies deploying AI to enhance existing locomotive and railcar monitoring systems. Progress Rail and Wabtec, the two largest U.S. manufacturers of diesel freight locomotives, have also developed AI-driven systems to help railroads and their onboard employees operate their locomotives as efficiently as possible, as well as proactively identify conditions that indicate potential maintenance issues before they become a major problem.  

Short lines own over 50,000 miles of track across the U.S.; maintaining and operating that network and ensuring compliance with Federal regulations requires significant investment of money and time. Short line railroads already use many different technologies in their businesses and draw data from many different sources. The unique ability of AI is to bring together and analyze large datasets to produce actionable insights that improve the efficacy and expand the capabilities of the technologies already in use by short line railroads.  

Companies such as Connixt and TekTracking offer solutions that help short lines, such as Finger Lakes Railway and the New Orleans Public Belt, record and organize information about the condition of their track infrastructure directly from the field. Analytic tools using AI also help prioritize investments to improve track conditions across their networks. Other industry initiatives, such as RailPulse, which is providing a foundation for widespread adoption of railcar telemetry solutions, are dramatically growing the pool of data that AI tools can access to improve short line performance. These systems will help railroads monitor real-time performance to help manage train handling and brake dynamics, as well as railcar asset health.  

AI enhancements of commonly used technology systems are helping improve operations and customer service. Transportation management system products from CedarAI and Wabtec have long helped short lines interface with both their customers and other railroads to manage shipments and settle transactions. Now the companies have both begun to integrate AI into their platforms to provide better estimates of when shipments will arrive. Providing rail customers with more accurate and up-to-date information on the location of their shipments, and when they are likely to arrive, has been a long-term goal in the rail industry. These improvements are expected to help rail more effectively compete for new business against other modes of transportation. This not only helps to grow business for short lines but also provides public benefits by moving freight off publicly funded and over-congested highways and onto private infrastructure that is the safest and most fuel-efficient option for moving freight over land. 

Short lines themselves are moving the ball forward on AI. At the American Short Line and Regional Railroad Association’s Annual Conference in April, railroad presenters from across the industry shared their progress in adopting generally available AI tools to streamline work across their businesses, including in areas that do not typically get as much attention as running trains, such as finance, HR, and IT. As the operators of a portion of America’s critical freight infrastructure, cybersecurity is a hot topic in the industry as short lines grapple with how to protect themselves from hostile nation-state actors and criminals in an environment when AI tools are both an increasingly powerful tool to attack but also defend technology systems. In addition to their own efforts, short lines are supported by companies like Cervello and Cylus, who have broad experience protecting large-scale critical infrastructure and are utilizing AI technologies to help apply that experience on short lines. 

With adoption of new technologies dependent on AI already changing the way short lines operate, railroaders are working with Congress and the Administration to ensure that they are positioned to continue providing critical and efficient rail service across the country. That includes pursuing robust funding for programs like Consolidated Rail Infrastructure and Safety Improvement grants that fund not only investments in physical short line infrastructure but can also be used to support technology initiatives and university research programs to drive AI innovation in rail. It is also crucial that Congress and Federal regulators allow flexibility for the rail industry to evaluate nascent rail technologies using AI and to deploy them rapidly after reasonable demonstration. As AI continues to develop and increase its capabilities, short lines will work on their own and with their vendor partners to adapt to a quickly evolving business environment. As railroads have done for almost 200 years, short lines will be innovative and resilient, seeking to safely apply the latest technology to provide efficient service to the tens of thousands of shippers who rely on rail to access markets and drive our economy forward. 

Fred Oelsner is the Vice President of Data, Technology, and Security, for the American Short Line and Regional Railroad Association