This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
October 2025
By Jeff Stagl, Managing Editor
For a long time, BNSF Railway has considered itself “a company of builders.” The Class I’s leaders use that characterization because throughout the company’s 176-year history, lots of infrastructure has been constructed, much rolling stock has been retooled and myriad new ways to perform tasks have been created.
Workers at the company and its predecessors have shared many of the same builder traits through the decades, and that continues today. More than 36,000 employees now help the railroad complete billions of dollars worth of infrastructure work, manage about 7,000 locomotives and thousands of rail cars, and transport nearly 10 million carloads and 10 million containers annually.
Overall, running the railroad is a massive undertaking that only a team of builders — who’ve developed a deep-rooted understanding of how the business works — can pull off, BNSF leaders say.
But in terms of technological advances, the railroad has not been a company of builders in recent years. Over the past decade, purchasing commercial off-the-shelf products has been the norm instead of developing technologies in house. And dozens of integration projects have incorporated purchased software.
So, BNSF leaders recently wondered: How do we get back to our roots technology-wise? The answer: bnsf | tech.
Created in March as a new company division, bnsf | tech is charged with developing technology solutions to address operational challenges, including various technologies designed to improve safety, on-time performance, terminal functions and the customer experience. The division emphasizes a customer-focused approach and aims to leverage internal expertise and collaboration.
Being more of a buyer than a builder with technology is antithetical to the railroad’s rich history, says Hari Govind, who joined BNSF in January 2025 as VP and chief technology officer and has spearheaded the Class I’s tech transformation.
“With bnsf | tech, we’re sort of reversing that tide. Slowly but surely, we’re becoming more builders than buyers,” he says. “I think we’ll be majority builders as we continue this transition.”
Among a plethora of projects, the new division is working to:
AI is a key component in bnsf | tech’s efforts. BNSF is “trying to bring AI to everything,” but at the same time be more focused on where it’s applied, says Govind, who served as chief information and technology officer for GEICO prior to joining the railroad and previously was senior VP of infrastructure and operations for Target Corp.’s tech team.
The Class I is hoping to redefine how employees can launch the company’s technology transformation via AI, said BNSF President and CEO Katie Farmer in emailed comments. At the railroad, the use of AI will always be about amplifying human judgment and creating new opportunities for the nation’s economy, she added.
“We feel we can define and refine that better than anyone else. We must lead all of our AI deployments with ethics, governance and trust,” Farmer said. “We are committed to focusing on outcomes with AI that go beyond experimentation and ties back to real value for us — operational excellence, cost savings and generating revenue, which in turn leads to safer operations, a better customer experience and smarter decisions for our nation.”
AI affords the railroad a big opportunity to transition from moving trains to becoming a freight mobility company, said BNSF Chief Data and AI Officer Anju Gupta in emailed comments. The railroad can be driven by the algorithmic optimization of its networks, routes, locomotives, fuel consumption and, most important, an accurate prediction of estimated times of arrival that will lead to on-time deliveries, said Gupta, who joined BNSF in June.
“This is about how we can change the freight economy of our nation and build end-to-end confidence within the railroad industry,” she said. “As we focus on transitioning ourselves to a differentiating freight mobility company, I am truly excited about a future where AI is the fabric of our enterprise, for the safety of our people, our cargo, our tracks and wheels, and everything in between.”
Deploying new technologies to better predict outcomes and find ways to work around potential problems or issues before they occur, such as weather events, is naturally going to improve the railroad’s performance, Farmer stressed.
Fewer service interruptions will improve safety, efficiency and resiliency. Using technology in conjunction with the workforce will enable the railroad to be proactive instead of reactive, Farmer said.
“Our vision with bnsf | tech is to create a space to not only improve our overall operations, but the global supply chain, and demonstrate the advantages of utilizing rail,” she said.
To meet its objectives, bnsf | tech is striving to deliver internal tech builds every two weeks. The division also has set a goal to provide predictable release schedules and complete incremental technological improvements every 90 days.
In the meantime, BNSF is trying to rapidly expand bnsf | tech’s workforce. Over the next year, the Class I plans to hire 1,000 employees. The railroad is actively recruiting computer science, data science and software engineers, particularly from colleges in various states, amidst a soft tech recession, says Govind.
“We just hired 100 engineers from state colleges in Wisconsin, Illinois, Michigan, Texas, Oklahoma and Florida,” he says.
The hiring market is interesting because there’s so much hype on AI and a lot of companies are making AI investments and recruiting, Govind says.
“This has been really fertile ground for us to come in and say, ‘Hey, we’re hiring,’” he says. “We have a lot of applicants for every job we’ve been posting.”
As more engineers join bnsf | tech, more projects will get underway, advance or reach completion. One of the division’s major undertakings is finding ways to improve efficiencies at terminals to boost throughput, increase container lifts, decrease overall truck movements and adjust sorting.
Truck drivers in a terminal can spend a half hour or more navigating the facility and trying to find the right parking spot for a container. By adding software to BNSF’s RailPASS mobile application for truck drivers, they can get turn-by-turn directions in a terminal and more quickly move a container to the right spot, says Govind.
“The app can tell them, here’s the ideal time for you to be there, here’s your time window and even to deliver to track side. It goes straight on the train and we’re off to the races,” he says. “When a train comes in for these containers to be picked up, we give them a pickup window. So, they spend the minimum amount of time in the terminal.”
More than 80,000 truck drivers currently use RailPASS — which can be downloaded onto Apple and Google smartphones — to expedite gate transactions, locate specific units and perform other essential in-terminal tasks.
Drones are factoring into container-location improvements, as well. Terminals in Los Angeles and Alliance, Texas, just north of Fort Worth currently are employing BNSF’s patented drone-based Automated Yard Check (AYC) system, which uses truck-mounted cameras, aerial footage and machine learning to triangulate the exact location of containers in 3D.
Workers at the terminals previously spent an entire shift trying to complete an inventory, but now they can maintain a real-time container inventory with the AYC system. To be used at most of BNSF’s 27 intermodal facilities by 2027, the system fosters faster turnarounds for truck drivers and improves service for customers.
Technologies such as RailPASS automated gate systems and AYC are worth the time, effort and dollars to develop, BNSF leaders believe.
“Those are the technologies we want to invest in because these things don’t really exist off the shelf, and where they do, are built to the lowest common denominator requiring loads of customization and integrations,” says Govind.
Another BNSF innovation that’s not available on any retail shelf: the patent-pending Load Plan Optimization (LPO) tool designed to help plan intermodal service. LPO uses AI to optimize how shipments are assigned to trains and routes.
Planning intermodal service is akin to a team relay race. By using an AI-based algorithm that can create a load plan for an outbound train in seconds, LPO helps ensure the right containers and trailers are located where they need to be on production tracks, minimizing the overall distance hostler drivers need to travel.
LPO and AYC supplement each other to help improve the customer experience and increase efficiency, BNSF leaders say.
The Class I also employs a patented Multi-Objective Train Block Assignment System to make switching operations less complex. The technology uses data and optimization models to assign train blocks to classification tracks in the most efficient method.
The system is designed to adapt to the current activity level in a yard, minimize unassigned volume, reduce switch moves and improve throughput. It can generate visual tools such as charts and diagrams to support real-time decision-making.
“Slowly but surely, we’re becoming more builders than buyers.” — Hari Govind, VP and chief technology officer
There are many other innovations under the bnsf | tech umbrella that incorporate AI. For example, AI solutions already enable the railroad to optimize throttle settings and coasting options based on various train features and factors, Gupta said in her comments.
“Fuel optimization for a greener future is key to how we will drive freight mobility in the years to come,” she said. “Fuel is one of our biggest expenses, and algorithmic optimization can be used to drive reductions and optimize fuel usage in how we run our network.”
The division also is working on predicting an accurate estimate of on-time performance for delivering freight to customers, in part to create a better customer experience.
“Driving an accurate demand forecast to real-time dynamic train scheduling with 100% capacity planning will be our future with a precise intersection of our data and AI journey,” Gupta said. “With our data and AI technology, we are excited about providing a future fluid network on our 32,500 miles of track enabled by optimal routing and yard optimization.”
Moreover, bnsf | tech is trying to enhance safety and predictive maintenance.
“AI accelerates us to a future where we have the ability to predict failure before it happens — and we are already doing that today,” said Gupta. “We are able to flag early signs of potential problems on the rails, on our wheels and with our signals as a natural part of our sequencing.”
Asset health is important to forecast preventative repairs and prevent the failure of parts, says Govind.
“This is one area where we’re making meaningful investments in AI,” he says. “We are exploring what AI algorithms can actually help us with these types of predictions.”
At the trackside, closed-circuit, high-resolution cameras are monitoring components while thermal imaging is helping to detect hot wheels. Other technologies are helping to identify cracked wheels or overloads or underloads based off the height of a shipment.
“These are examples of where we’re using vision computing and we’re using AI models with vision computing,” says Govind.
For the engineering and mechanical departments, bnsf | tech is helping to transition from better diagnostic software to more prognostic software for parts and components. The effort can help with preventative repairs, preventative swaps, tracking parts’ usage and lifespan, and keeping tabs on environmental conditions. Predictions can be baked into AI models, says Govind.
There’s no reason why all locomotive and rail-car parts can’t be tracked, such as when they were installed, how long they last or how much wear and tear they received, he says. Based off the collective history, BNSF then could predict when a part might fail.
“The things that cause derailments and the things that cause unscheduled stoppages, that has a massive ripple impact across the network,” says Govind. “The sooner you know something might fail, the quicker you can respond.”
Weather has a massive impact on the network, too. It’s the single largest operational disrupter. Weather patterns are getting more extreme and although there are good prediction models available, they’re not 100% accurate, or even 90% accurate, which would provide a better response time, says Govind.
Recently, BNSF began placing sensors near track and on trains that provide the wind speed in a region and the wind’s direction.
In some areas in the Plains, wind speeds at times can reach 50 mph, with gusts up to 70 mph, so trains need to stop.
“Now, we’re stopping 30% fewer trains because our sensors tell us it’s either a headwind or a tailwind,” says Govind. “It’s actually safe to move the train. You only want to stop it if it’s a crosswind.”
As bnsf | tech continues to evolve, one thing Govind plans to address is the reputation of some technologists as being disconnected from the problems they’re trying to solve.
“What’s going to be better is every engineer who can connect the dots between the work that they’re doing and the impact it has on the railroad,” says Govind.
For example, if there’s a loose cable in a data center, a network engineer needs to understand that a simple loose connection means software won’t run and then there’s many wasted opportunities at a terminal, he explains.
“I think our engineers are starting to understand they’re not just a network engineer, a data engineer or a data scientist. Everything they do touches everything that our operators are doing and directly impacts our customers,” says Govind.
What will help the Class I complete its tech transformation is if every employee understands there are too many valuable opportunities available that can be utilized to make the railroad better in every facet of what it does, said Farmer.
“We must continue to build, look for new ways to improve and grow, and meet the ever-changing economic landscape and needs of our customers,” she said. “We are laser-focused on our customers, building better ways to manage and track their fleet, schedule work at their facilities, improve their overall experience and build more predictability into our network.”
Email questions or comments to jeff.stagl@tradepress.com.
Related Topics: