How UPS Saved 100 Million Miles With One Operations Rule

Lucas: So there's this one rule at UPS that sounds almost like a joke: drivers are told to avoid left-hand turns. No lefts. Only rights. Luna: I've heard that - it's one of those 'too weird to be true' operations stories.

Lucas: Well, it's true. And it's been one of the most consequential single rules in logistics history. UPS estimates it saves them about 100 million miles of driving per year - which translates to 10 million gallons of fuel and roughly 100,000 metric tons of CO2. Luna: Hundred million miles - just from not turning left?

Lucas: That's the number. And the rule has been in place since 2004, though its origins go back to the 1960s. A driver named Don Howery noticed that sitting in left-turn lanes, waiting for a gap in oncoming traffic, was burning time and fuel. So he started planning routes that only turned right.

Luna: That's almost 20 years before the rule went company-wide. What took so long? Lucas: Good question. It became a kind of underground best practice among veteran drivers.

But UPS didn't formalize it until they started building route optimization software - which they call ORION - in the early 2000s. ORION stands for 'On-Road Integrated Optimization and Navigation.' And when they fed the data in, the algorithm kept recommending right-turn-heavy routes. That's when they realized the drivers had been onto something for decades.

Luna: So the software confirmed what the drivers already knew. That's actually a nice story - the company listening to its people. Lucas: It is - up to a point. The reality is more complicated.

ORION was supposed to replace driver judgment entirely. But UPS found that drivers resisted the system because it didn't account for local knowledge - like which alleys are blocked on trash day, or where construction is happening. Luna: So they had to blend the two - algorithm plus human. Lucas: Exactly.

And that tension is the real operational story here. The no-left-turn rule is simple enough to remember. But implementing it at scale, across 55,000 routes in the US, required UPS to wrestle with something every operations manager faces: the gap between a perfect theoretical process and the messy reality of execution. Luna: Let's stay on the numbers for a second.

A hundred million miles saved - how do we know that's real? Lucas: UPS publishes it in their sustainability reports. They've been tracking it since ORION was deployed. In 2016, they claimed the system saved 100 million miles and 10 million gallons of fuel.

Those numbers have stayed pretty consistent year over year. Independent analysts have confirmed the methodology - it's based on GPS data from every truck. Luna: But wait - a right-turn-only route means you're sometimes going out of the way. How does that save miles?

Lucas: That's the counterintuitive part. A right-turn-only route is often physically longer in distance. But the key metric isn't distance - it's time. Left turns involve waiting for a gap in traffic, and that wait averages 15 to 30 seconds per turn.

Multiply that by 120 to 160 stops per driver per day, and the time adds up fast. Plus, idling while waiting burns fuel. So a route with more turns but zero waiting time is actually faster and more fuel-efficient. Luna: So the rule isn't about shortest distance - it's about minimizing delay.

Lucas: Right. It's a classic operations insight: the bottleneck isn't the road, it's the intersection. And by eliminating left turns, UPS reduced the time drivers spend in the most delay-prone part of their route. The rule also reduces accident risk - left turns are statistically more dangerous than right turns.

Luna: I've seen that stat. Left turns account for something like 60 percent of intersection crashes, even though they're a smaller share of total turns. Lucas: Exactly. So safety was a secondary benefit.

Fewer accidents mean fewer insurance claims, less vehicle damage, less downtime. UPS doesn't break out those savings publicly, but analysts estimate they're in the tens of millions annually. Luna: Let's talk about the limits of this rule. Does it apply everywhere?

Urban vs rural? Lucas: It's primarily optimized for urban and suburban routes - where there are intersections and traffic lights. On rural routes with minimal traffic, the benefit is smaller, and drivers have more discretion. The ORION system actually relaxes the rule when it determines that a left turn wouldn't cost much time.

Luna: So it's not a rigid rule - it's a parameter in an algorithm. Lucas: Exactly. And that's the smart part. A blanket 'no left turns' edict would fail.

But a system that biases toward right turns, and only allows a left when the algorithm calculates it's faster - that's operationally sound. Luna: Has any other company tried this? Lucas: A few. FedEx has a similar preference but doesn't enforce it as strictly.

Some municipal garbage trucks use right-turn-only routes. But the most famous failed attempt was Domino's Pizza. In the early 2000s, they tried to mandate no left turns for delivery drivers. It backfired because pizza delivery routes are short and unpredictable - drivers were making illegal U-turns and getting lost.

The rule was quietly dropped. Luna: So context matters. A rule that works for a 10-hour delivery route doesn't work for a 30-minute pizza run. Lucas: Exactly.

And that's the broader lesson: operations rules need to be matched to the specific operational context. UPS drivers have fixed routes they repeat daily. A pizza driver's route changes every delivery. Luna: So the no-left-turn rule is really a story about understanding your bottlenecks.

Lucas: It is. And it's a story about the relationship between software optimization and human judgment. ORION was originally designed to automate route planning entirely. But UPS found that when they overrode driver input, performance actually dropped.

Drivers would game the system - deliberately driving inefficiently to prove the software wrong. Luna: Oh, that's classic. The Hawthorne effect in reverse. Lucas: Exactly.

So UPS had to redesign ORION to be a decision-support tool, not a decision-making tool. The algorithm recommends a sequence of stops and a turn pattern, but the driver can override it for specific local reasons. That hybrid model improved compliance and performance. Luna: It's almost like the algorithm needs to earn the driver's trust.

Lucas: And that trust builds over time. UPS found that new drivers accepted ORION more readily than veterans. But after about six months, even the skeptics started seeing the logic - especially when they compared their fuel consumption to colleagues who followed the system. Luna: I want to zoom out a bit.

UPS spends about $1 billion a year on fuel. So saving 10 million gallons - at roughly $3 per gallon - is about $30 million in direct savings. That's real money, but it's less than 0.1 percent of their revenue.

Lucas: That's the right frame. On its own, the fuel savings are modest for a company with $100 billion in revenue. But the real value is in the cumulative effect. The rule also saves time - and time is capacity.

UPS estimates the no-left-turn rule adds about 5 to 7 percent more stops per driver per day. That's huge. More packages delivered without adding trucks or drivers. Luna: So the operational leverage is massive.

A small change in routing compounds into millions of extra delivery capacity. Lucas: Exactly. And that's why the rule has been so durable. It's not a gimmick - it's a marginal gain that adds up at scale.

Every second saved per stop, multiplied by 55,000 drivers and 250 days a year, is a significant number. Luna: Before we wrap - has UPS ever considered eliminating left turns permanently? Like, mandated no exceptions? Lucas: They considered it, but the data showed that about 10 percent of the time, a left turn is genuinely faster or necessary.

So they kept it as a strong preference, not an absolute. That flexibility is probably why the rule has lasted 20 years. Lucas: You know, before we wrap up - and I don't want to break the flow here - but a quick honest thing. This show is possible because a small group of listeners chips in monthly through buy me a coffee dot com slash fexingo.

That's it - no ads, no sponsors, just that. Luna: Yeah, and it's a pretty modest group. But it's enough to keep us making these deep dives every week. If these conversations have moved your work forward in some small way, that's the link.

Lucas: And we mean that genuinely. It's not a guilt trip. We just wanted to mention it because the economics of a niche podcast like this are simple - either listeners fund it, or it doesn't exist. So if you've gotten value, that's the door.

Luna: Alright, back to the route. So the no-left-turn rule - it's almost twenty years old now. Is it still relevant in the age of real-time traffic data and AI? Lucas: More relevant than ever.

UPS is now using machine learning to predict traffic patterns and adjust routes in real time. But the right-turn bias is still a core parameter in the algorithm. The principle - minimize time spent waiting at intersections - is timeless. Luna: So the rule is less about the specific maneuver and more about a mindset: identify the biggest source of delay in your process, and eliminate it systematically.

Lucas: Exactly. The no-left-turn rule is just one example. The real operations lesson is to watch where your people spend time waiting, and find a way to design that wait out. For UPS, it was intersections.

For a hospital, it might be patient handoffs. For a factory, it might be machine setup time. The specific fix will differ, but the approach is the same. Luna: And it always comes back to listening to the people doing the work.

Lucas: That's the part I find most compelling. The rule started with a driver's observation. The software confirmed it. But the implementation only succeeded because UPS respected the driver's autonomy to override when needed.

The algorithm alone couldn't do it. Luna: So the next time you see a UPS truck make three right turns to go around the block, you'll know why. Lucas: And you'll know they're not going in circles - they're optimizing.