Latest / The Pivot Podcast with Fexingo: How Businesses Change Direction and Survive Disruption / How a Railroad Pivoted Into a Data-Driven Logistics Giant
Transcript
- Lucas: If I told you that one of the most sophisticated data operations in the United States runs on 32,000 miles of steel laid down in the 1860s, you'd probably think I was talking about a telecom company or maybe a cloud provider. Luna: But you're talking about a railroad. Lucas: Union Pacific, specifically. The company was founded in 1862, finished the transcontinental railroad in 1869. And for most of its history, it did exactly what you'd expect: haul coal, grain, cars, containers from point A to point B. But starting around 2018, something shifted. The company quietly pivoted from being a freight hauler to being a data-driven logistics platform. Luna: I remember reading about precision scheduled railroading in the mid-2010s — that was the big trend. Run trains like a Swiss watch, maximum efficiency, minimum dwell time. But Union Pacific had a pretty rough time with it. Lucas: They did. The model was pioneered by a guy named Hunter Harrison at Canadian National and then CP Rail. The idea is simple: run fewer, longer trains on a fixed schedule, and squeeze every ounce of asset utilization. Wall Street loved it. But Union Pacific's network is enormous and diverse — they have single-track lines through the Rockies, they deal with extreme weather, and they serve customers who don't always have their freight ready at 2 PM on a Tuesday. The one-size-fits-all schedule caused massive congestion. Luna: So what did they do? Abandon precision scheduled railroading altogether? Lucas: Not exactly. They kept the core principle — run scheduled trains — but they layered a massive data infrastructure on top. CEO Lance Fritz said publicly that the railroad needed to be 'predictable, not just efficient.' So they started building what they call the 'Unified Plan 2030' — a real-time operations platform that uses machine learning to predict where congestion will hit and adjust schedules dynamically. Luna: Can you give me a concrete example of how that works on the ground? Lucas: Sure. Every train on Union Pacific has GPS tracking, and every locomotive sends telemetry data — speed, fuel consumption, brake pressure, engine temperature — every few seconds. That data feeds into a model that predicts train meets. In a single-track territory, if two trains are heading toward each other, the system calculates the optimal meeting point to minimize wait time. It used to be done by human dispatchers with radio calls and gut feel. Now it's done by an algorithm that reassigns priorities in real time. Luna: And the results? I imagine the fuel savings alone would be significant. Lucas: Union Pacific reported that by 2023, they had reduced fuel consumption by six percent compared to 2019 levels, while improving on-time delivery by twelve percent. On a network that burns about one point two billion gallons of diesel a year, six percent is enormous. That's roughly 72 million gallons saved. And the on-time improvement matters because their customers — companies like UPS, Amazon, and auto manufacturers — are running just-in-time supply chains. A delayed train can idle a factory. Luna: So the pivot really was from an asset-centric view — how many trains, how many cars, how fast — to a customer-centric view. What does the customer need, and how do we deliver it reliably? Lucas: Exactly. And the data infrastructure is what enabled that shift. They built something called the 'Network Operations Center' in Omaha, which looks more like a NASA mission control than a railroad dispatch. There are dozens of screens showing real-time train positions, weather overlays, crew availability, and even predictive maintenance alerts. They can see a problem forming three hours before it happens and reroute traffic. Luna: I want to ask about the crew side of this. Railroads have a famously unionized workforce, and precision scheduled railroading was notorious for burning out crews. How did they manage that transition? Lucas: That's a great question. The old model expected crews to be on call 24/7 and sometimes they'd sit in a hotel for days waiting for a train. The new system uses the same telemetry data to predict crew demand. They now have a 'crew optimizer' that schedules engineers and conductors based on actual train movements, not guesses. The result is fewer wasted hours and more predictable work schedules. The union — the Brotherhood of Locomotive Engineers and Trainmen — has been cautiously supportive because it reduces the worst part of the job: the uncertainty. Luna: So it's a rare case where operational efficiency and worker quality of life improved together. Lucas: Yes, and that's part of why this pivot is interesting beyond just the railroad industry. Every old-economy company — utilities, trucking, manufacturing — is facing pressure to digitize. But the playbook isn't 'rip out your legacy systems and become a tech company.' It's more like: use data to make your existing operations smarter. Union Pacific didn't stop running trains. They just started running them with better information. Luna: And the economics of that are compelling. Rail is already the most fuel-efficient way to move freight over land — one train can carry the load of several hundred trucks. If you make it even more efficient, you're not just saving money, you're reducing carbon emissions significantly. Lucas: Right. Union Pacific estimates that each train removes about 300 trucks from the highway. So every incremental improvement in rail efficiency has a compounding environmental benefit. That's part of why the company's sustainability report now highlights data-driven initiatives as a core pillar of their carbon reduction strategy. Luna: I'm curious about the competitive landscape. How does Union Pacific compare to other Class I railroads like BNSF or Norfolk Southern on this front? Lucas: BNSF, which is owned by Berkshire Hathaway, has also invested heavily in technology — they have a similar operations center in Fort Worth. Norfolk Southern had a derailment in East Palestine, Ohio in 2023 that really shook public trust, and they're now playing catch-up on safety analytics. Union Pacific has arguably been the most aggressive in using machine learning for predictive maintenance. They inspect tracks with drones and computer vision, and they use acoustic sensors to detect wheel bearing defects before they cause a derailment. Luna: That feels like a different kind of pivot — from reactive safety to proactive safety. Lucas: Absolutely. And it's worth noting that this wasn't a sudden pivot. It took about five years from the initial investment in data infrastructure to seeing measurable results. Lance Fritz started the transformation in 2018, and they only fully deployed the Unified Plan in 2023 after a lot of trial and error. That's a patient capital story that doesn't always get told. Luna: Speaking of patient capital — and I know this is a bit off the beaten track — but it strikes me that shows like this one, which dig into how companies actually change, are only possible because of listener support. We don't run ads, and we don't have a big corporate backer. It's just a handful of people who chip in through buy me a coffee dot com slash fexingo, and that literally covers the production costs for episodes like this. Lucas: Yeah, it's a small group, but it makes a huge difference. And we're grateful. It keeps the show independent and lets us pick stories like this one that might not get mainstream coverage. Luna: Anyway, back to the railroad. Lucas, do you think this data-driven model is replicable for other industries, or is Union Pacific a special case? Lucas: I think it's replicable, but it requires two things: a willingness to invest in data infrastructure without expecting immediate ROI, and a leadership team that understands the operational detail. Union Pacific's CEO came from the operations side. He knew how trains actually run. That's more important than hiring a Silicon Valley data scientist who doesn't understand the physics of a 10,000-ton train going downhill. Luna: So the lesson is: don't digitize for digitization's sake. Digitize to solve a specific operational pain point. Lucas: Exactly. And measure what matters. Union Pacific didn't just track 'data points.' They tracked fuel consumption, on-time percentage, crew utilization, and derailment rates. Those are the metrics that connect directly to business outcomes. Any company thinking about a pivot can learn from that discipline. Luna: And maybe the biggest takeaway is that pivoting doesn't mean abandoning your core. It means making your core smarter. Lucas: Well said. Union Pacific is still a railroad. It still hauls freight on steel wheels. But the way it does it has changed fundamentally. And that's what survival looks like in an industry that's been around since the Civil War.