The role of AI in short line operations is a hot topic at ASLRRA event – RailPrime

Bridget Dean, Senior Associate Editor

Artificial intelligence (AI) and self-driving technology have been a hot topic in the rail industry for the past few years. And the use case for AI tools and systems in short-line operations is strong, according to speakers at the American Short Line and Regional Railroad Association’s annual conference and trade show, held April 12-14 in Minneapolis.

Two breakout sessions covered this topic. One focused on AI for sustainability applications, and the other looked at how to incorporate AI into everyday workflows in short lines. Speakers in both sessions emphasized how AI systems and tools can structure underutilized data to make information more accessible, improve workflows, and make operations more efficient.

AI Use Cases in Short Lines Many AI tools in use today combine several types of AI systems, including machine learning, deep learning, natural language processing, computer vision, and generative AI. Generative AI is a well-known type of AI system used in programs such as ChatGPT and other “chat bots” or image generation systems. said Jason Smeek, vice president of global supply chain, logistics and industrial development at AECOM, who hosted the AI for Sustainability session.

Railroads are already using AI systems for predictive maintenance, emissions measurement, energy management, regulatory compliance, environmental resiliency, hazardous materials response, and more, Smeek shared in a short introductory presentation. The session then moved to three panelists, each of whom introduced their AI service offerings and demonstrated use cases for AI in short-haul operations.

Dustin Bullard, CEO of Positive Train Control Services LLC, introduced one of his products, the Locomotive Intuitive Support Assistant (LISA4). The LISA platform is designed to reduce locomotive downtime by helping technicians troubleshoot failures, Bullard said.

The platform organizes technical knowledge from locomotive schematics, analyzes failure symptoms, narrows down possible failure points, and guides engineers to the best next steps to find the root cause. Bullard emphasized that LISA eliminates wasted time, unnecessary material use, and labor associated with a trial-and-error approach to maintenance.

“LISA can model a locomotive’s electrical system and predict how faults will behave in real-world situations,” he said, adding that the platform does not replace professional engineers, but rather allows access to information when one is not available.

The second panelist, Beacon Advisor Simon Davidoff, shared how the company’s AI Traffic Management Center-React (TMC-React) platform was recently implemented into the Georgia Department of Transportation’s incident response dispatch system. The TMC-React system, like LISA, allows people to quickly get the information they need without having to rely on external support or having to dig through hundreds of paper files or digital folders.

TMC-React takes your internal data such as SOPs, contacts, and construction documents to create a searchable and chattable database. All answers on the platform link to the source document, so users can trust the answers, Davidoff said. Although the system was built for the state Department of Transportation, the core technology could also be implemented for short-haul operations, he added.

RailState’s Dan Devoe (pictured) was one of three panelists for the “AI for Sustainability” breakout session.bridget dean

The third panelist, Dan DeVoe, director of marketing for RailState, pitched the company’s trackside imaging system as a tool for short lines to independently verify vehicle movement, product trends and freight exchanges without relying on self-reporting from partner railroads.

RailState’s proprietary sensors are installed outside railroad rights-of-way throughout the United States. They continually collect identification information, train speeds and hazardous materials placard information on rail cars and containers, DeVoe said.

That information is automatically uploaded to RailState’s cloud, where computer vision and agent AI systems classify the data to derive big-picture datasets for train traffic, product flow, congestion points, and more. DeVoe said users can access that data online within 30 minutes of a train being scanned. For short routes, this means time-stamped verification of freight flows without relying on data from multiple partners or railways.

Although the three panelists each had different use cases for AI systems, each presentation hit on the same key points. Railways frequently use data to make decisions, and AI systems can collect and structure that data more efficiently, leading to more sustainable operations.

Subheading: Get started with AI

The second AI session featured presentations from Simpletech Innovations, OmniTRAX Inc., Alaska Railroad Corp., and Wi-Tronix covering a variety of use cases across short rail lines.

For example, Wi-Tronix’s Violet platform allows short-line operators to “see, hear, feel and experience” what’s happening inside and outside the cab of a locomotive from miles away, said Lisa Matta, the company’s chief innovation officer. Ritu Chawla, director of AI platform development at Wi-Tronix, added that the system can detect phone usage in taxis, identify trespass hotspots, verify radio calls, and monitor operational data such as fuel and brake activity.

OmniTRAX also uses AI technology on locomotives to compile data, said Leah Twombly, director of enterprise services for the railroad and real estate holding company, and Beth Fleischer, Alaska Railroad’s director of IT, added that AI has proven useful in office administrative tasks. Shortline had no AI policy and had security concerns until Fleischer learned that employees were using AI tools for work purposes without guidelines.

“If you think people aren’t using AI in the office, you’re wrong,” Fleischer said. “Learn what they’re doing and make sure it’s done correctly.”

Overall, there are many use cases for AI systems in the short delivery industry, but there can be a steep learning curve for those new to the technology. Andrew Hollister, CEO of SimpleTech, left session attendees with some advice to get started. Here are some key points:

Different AI tools are suitable for different tasks. Go beyond ChatGPT and find a tool that better understands what you need. For coding, try Claude. Try NotebookLM for research.

Your company may already have an AI policy in place. Before you start using AI tools in your business, know what they are.

If you are not paying for the product, you are the product. Large language models such as ChatGPT are trained on large amounts of data, including the content you upload. If possible, subscribe to the AI tool of your choice and turn off data sharing and model training. Otherwise, there is a risk that non-public company information shared with AI tools will be leaked back to the public.