Latest / The Operations Podcast with Fexingo: Process, People, and Profit in Modern Business / How a Bus Fleet Cut Fuel Costs by 25 Percent with Route Optimization
Transcript
- Lucas: So there's this midsize transit authority in Fort Worth, Texas, called Trinity Metro — runs about 160 buses across the city. And last year they did something that sounds almost too simple: they changed the way their buses were routed. Not the physical streets, the logic of which bus goes where and when. And it cut their fuel bill by 25 percent. Luna: Twenty-five percent on fuel alone? That's huge. What was the old system? Lucas: Old system was basically static. Routes were drawn years ago based on census data and manual observations. Dispatchers had some flexibility but mostly it was 'bus A runs route B at time C' every single day. No adjustment for real-time demand, no feedback loop. So you'd have buses running nearly empty on some routes while others were packed. And a lot of idle time — buses sitting at terminals with engines running, waiting for the next scheduled departure. Luna: Idle time is brutal. Buses get maybe four to six miles per gallon. Sitting still burning diesel is just throwing money away. Lucas: Exactly. So Trinity Metro partnered with a startup called Routal — they do optimization software for fleet operations. And they deployed a dynamic routing algorithm that ingests three streams of data: real-time passenger count via onboard sensors, traffic data from city cameras and GPS, and the status of each bus — fuel level, maintenance schedule, even charge level for their six electric buses. Luna: Right, because if you have an electric bus in the fleet, you can't just swap routes arbitrarily — it needs to stay within range of its charging station. Lucas: Exactly. The algorithm solves a constrained optimization problem every 15 minutes. It reassigns buses to routes to minimize total distance traveled, idle time, and fuel consumption, subject to all those constraints — range, driver shift limits, regulatory rest periods. And it pushes revised schedules to the drivers' tablets. Luna: And the drivers actually followed the new schedules? That's often the hardest part. Lucas: You're right — the human factor. The rollout had its bumps. Drivers were used to their regular routes. Suddenly they're told at 10 a.m. that they're switching to a different line. Some pushed back. So Trinity Metro did something smart: they ran a pilot with a small group of volunteer drivers first, gave them a bonus for participating, and used that group's feedback to tweak the system. Then they rolled it out gradually over four months. Luna: Plus they had to integrate with the existing scheduling software, which was probably from the 1990s. Lucas: Oh, it was. Their legacy system was a terminal-based platform called Trapeze — it's actually pretty standard in transit, but it wasn't built for real-time optimization. Routal built an API layer that pulled data out of Trapeze and sent new schedules back in. That integration alone took about six weeks. Luna: So what were the actual numbers after the system went live? Lucas: Over the first twelve months, Trinity Metro reported a 25 percent reduction in fuel consumption — from about 1.1 million gallons of diesel to roughly 825,000. That's about $1.2 million in savings at current diesel prices. Idle time dropped 40 percent. And passenger wait times actually decreased by about 12 percent on average, because buses were better matched to demand. Luna: So it's a triple win: lower costs, less emissions, better service. And all from changing software, not buying new buses. Lucas: Right. The capital outlay was modest — about $400,000 for the software implementation and the onboard passenger sensors. So they recouped that in about four months. Now they're looking at expanding the system to include paratransit vehicles — those are the smaller vans for disabled riders — and eventually coordinate with the local school district's bus fleet. Luna: If today was actually useful to you, the way these stay ad-free is listener support — buy me a coffee dot com slash fexingo. Lucas: Yeah, it's a small gesture that keeps the podcast independent and means we can keep digging into case studies like this one. So thank you to those who already do. Luna: Alright, back to Trinity Metro. What about the electric buses? Did the algorithm handle them differently? Lucas: Great question. The six electric buses have a range of about 150 miles in ideal conditions, but in Texas summer, with AC running, that drops to maybe 110. So the algorithm treats them as 'range-constrained' assets. It assigns them to shorter routes and ensures they return to the depot before hitting a 20 percent charge threshold. That's a hard floor — the system won't dispatch a bus below that level. Luna: And does the algorithm account for charging time? Because those buses take hours to recharge. Lucas: Yes. It schedules charging during off-peak periods when electricity is cheaper, and it staggers the charging so not all six are drawing power at once — that would blow the depot's transformer. So there's a sub-optimization for the charging schedule as well. Trinity Metro estimates they saved another $30,000 a year on electricity costs just from that load management. Luna: That's the kind of granular detail I love. Now, one thing I'm wondering: does this algorithm work for other fleets? Like delivery trucks or ride-share? Lucas: The core logic is generalizable. Routal has since deployed similar systems for a package delivery company in Dallas and a food distributor in Chicago. The key is the quality of the real-time data input. If your passenger count sensors are inaccurate, or your GPS data is spotty, the optimization degrades fast. Trinity Metro invested in sonic passenger counters at each door — they're about 98 percent accurate. Luna: So it's not plug and play. You need the data infrastructure first. Lucas: Exactly. That's the prerequisite. But once you have it, the ROI can be compelling. And it's not just fuel. Trinity Metro also saw a reduction in maintenance costs because buses were running fewer miles overall — about 8 percent fewer miles across the fleet. Less wear on brakes, tires, transmissions. They estimate an additional $200,000 in annual maintenance savings. Luna: So total annual savings around $1.4 million on a $400,000 investment. That's a 3.5x return in the first year. Lucas: Right. And the system is still improving. The algorithm learns from historical data — it gets better at predicting demand on holidays, during special events, even during weather disruptions. For example, it learned that when the local college has a football game, certain routes see a 300 percent spike in ridership. Now it pre-positions extra buses near the stadium. Luna: That's the kind of adaptive behavior you can't get with static schedules. You'd need a really experienced dispatcher to know that, and even then they might forget. Lucas: Exactly. And the algorithm doesn't forget. It also doesn't get tired or distracted. So the role of the dispatcher shifts from making minute-by-minute decisions to exception handling — dealing with accidents, breakdowns, that kind of thing. The frontline managers at Trinity Metro told me their dispatchers feel more empowered now because they can focus on the unusual events instead of routine scheduling. Luna: What about the drivers? Did they eventually come around? Lucas: Mostly, yes. After the pilot and phased rollout, the drivers saw that the system wasn't arbitrary — it made logical sense. And they got a small productivity bonus for hitting on-time performance targets, which improved. Driver turnover actually dropped 15 percent in the first year. The transit authority attributes that to less frustration with outdated schedules and more consistent routes. Luna: That's an unexpected benefit. So the operations improvement rippled into HR. Lucas: Exactly. And it's a good reminder that operational changes often have second-order effects that you don't anticipate. The CFO originally signed off on the project purely for fuel savings, but now they're seeing benefits in maintenance, driver retention, and customer satisfaction. The on-time performance improved from 78 percent to 91 percent. Luna: So what's the next frontier for this kind of optimization? Is it integrating with traffic signals? Lucas: That's exactly what they're piloting now. The city of Fort Worth is working with Trinity Metro on a bus priority system at traffic lights — so when a bus is running a few minutes behind, the signal stays green a bit longer or the red shortens. That requires the traffic management system to talk to the routing algorithm. Early tests show it could cut trip times by another 5 to 10 percent. Luna: And that would save even more fuel, since less time idling at red lights. Lucas: Exactly. And it reduces brake wear too. So the whole system becomes more efficient. The interesting thing is that this isn't about any single breakthrough technology. It's about connecting existing systems — passenger counters, GPS, traffic signals, charging infrastructure — with a smart optimization layer. Trinity Metro didn't invent anything new; they just got the data to work together. Luna: Which is honestly harder than inventing something new. Integration is the messiest part of operations. Lucas: It is. And that's why so many organizations get stuck. They buy a best-in-class routing tool but it doesn't talk to their payroll system, or the fuel card data is in a separate silo. Trinity Metro's IT director told me the biggest challenge was not the algorithm — it was convincing six different vendors to share data on a common platform. Luna: So there's a lesson there for any operations manager: before you buy the flashy optimization tool, make sure your data pipes are clean. Lucas: Absolutely. Clean data, consistent APIs, and a willingness to change processes. The technology is not the hard part anymore; the organizational alignment is. Luna: Alright, I think that's a good place to leave it. Next time we'll look at how a hospital applied similar logic to operating room scheduling. Lucas: That's a great one. Operating rooms are like buses — they're expensive assets that need to be utilized efficiently. See you next time.