One truck, one trailer, eight driving days, Arizona to Wisconsin and back. Every second of it recorded. Here is what worked, what it cost, and what we still can't tell you.
A Ford F-150 PowerBoost hybrid towed a LightShip AE.1 travel trailer from Prescott, Arizona to Oshkosh, Wisconsin and home again — 3,719 miles through the Rockies, across the plains, and back over the Colorado Plateau in the middle of July. The two vehicles were connected by a charging cable, so the truck could refill the trailer's battery from its engine while driving down the highway.
Nothing about the trip was staged. It was a real trip with a real destination, in real July heat, at normal highway speeds. The difference is that the truck was wired to a data logger recording forty-two measurements every two seconds, plus a separate GPS logging position once a second. That comes to about 1.3 million readings.
The point was to answer three questions honestly, with numbers instead of brochure claims:
This part works, and it works consistently. The truck's onboard generator sent 5.8 kilowatts into the trailer's battery whenever it was running — on day one, on day eight, in the mountains, on the plains, at 60°F and at 100°F. The rate never sagged. That was the single most reliable finding of the whole trip.
What it costs is about three-quarters of a gallon of gas per hour of charging. We measured that two completely different ways — once by comparing a matched round trip in June, and again by statistically separating the charging load out of 18,500 highway measurements from this trip. The two methods, which share no assumptions, landed within a few percent of each other. That is about as good as field measurement gets.
Put another way: five hours of driving with the charger on puts roughly a third of the trailer's battery back, and costs about three and a half gallons. Filling the pack from empty would take about fourteen hours of driving and ten and a half gallons.
Towing without charging, the rig returned 15.3 miles per gallon. That figure is unusually trustworthy — we got it twice, on two different days, over 1,080 miles, in weather 15°F apart, and the two results differed by one percent.
Turn the charger on and hold it on, and you should expect about 13 mpg at highway speed. Across the whole trip, charging accounted for 3.8% of all the fuel burned — under ten gallons out of 259 — because the charger wasn't running most of the time.
Here is the finding a brochure would leave out, and it is not the one we thought we were writing. The charging rate never dropped. What dropped was how much we were willing to run it as the day got hot — and that was our decision, not the equipment's.
One day in Nebraska shows it cleanly, because we drove three segments in a row as the afternoon heated up:
The obvious reading of that chart is that the equipment gives up in the heat. That is not what happened. The driver deliberately cut the charger back as the ambient climbed into the 90s, to keep engine coolant temperature in hand — the truck is already working hard pulling 15,000 pounds, and running the generator on top of that adds heat exactly where there is least margin for it. Nothing in the recorded data separates a driver's decision from an equipment limit; the two produce identical-looking traces. We know which one this was because it was written down at the time, and that is why every charger state change now gets its reason logged as it happens.
Whether the charging hardware has a hot-weather ceiling of its own is genuinely open. We hit exactly one temperature-related stop in 3,719 miles — a connector on a Colorado climb, which started again by itself as the air cooled. One event does not establish a threshold, and the figure we began the trip with, that it quits somewhere in the low 100s, was an estimate we never got to verify. That same event took 30 to 45 minutes to restart, well after everything had cooled. We can say with reasonable confidence the delay was not on the truck's side, because the truck's own cutoff sits about 100°F above where this happened. How long that delay usually runs, we have seen once. Once is not a number.
Plan around itCharge in the morning — not because the system quits in the afternoon, but because that is when you have engine thermal headroom to spend, and because arriving at camp with a full battery matters more than it sounds. An occupied trailer overnight in summer draws one to two kilowatt-hours an hour just for air conditioning.
With the cable connected but the trailer's battery not actually taking a charge, the system still moves about a kilowatt. Our first pass through the data called that overhead — a kilowatt burned for nothing — and recommended switching it off between charging sessions. That was wrong, and it is worth saying so plainly rather than quietly deleting it.
A kilowatt is very close to what the trailer draws simply existing: fridge, twelve-volt system, fans, controls. The truck isn't wasting that power, it's covering those loads directly, so the trailer's battery drains more slowly than it otherwise would. It doesn't push the state of charge up, which is why it looked like nothing was happening. Over the trip about 16 kilowatt-hours went across the cable that way, costing roughly two gallons. Switching it off, as we originally advised, would just have moved that load back onto the trailer's own battery.
EstimatedThe kilowatt itself is measured. What the trailer does with it is our reading, not our measurement — the truck's logger sees what leaves the truck and nothing of what happens at the other end of the cable. Settling it takes about an hour on the next trip: watch the trailer's battery drain with the cable plugged in, then unplugged, same conditions, and compare the two rates.
The trailer has a motor in its own axle. Switch it on and it takes over some of the work of moving itself, and the truck's fuel consumption drops immediately and obviously.
We ran a careful test outside Denver: the same 35-mile loop twice, back to back, once with the trailer motor off and once with it on. Same road, same traffic, elevation within two feet at the end.
The raw fuel rate fell 47%. That number is real but it is not the honest one, because the two systems can't run at the same time — switching the trailer motor on also switches the charger off, so part of that 47% is simply the charger no longer running. Separating the two effects, the trailer motor's actual contribution is a 27.8% reduction, distance for distance. Still a substantial number, and still the largest single efficiency lever in the system.
The catchThe trailer spent more battery than the truck's charger could put back — by a wide margin, under every reasonable way of valuing the energy. Used as a fuel-saving device on a trip with no plug at the other end, it loses. Used on a trip where you plug in overnight, the arithmetic flips and it wins, because the energy it spends is cheap grid power rather than gasoline run through a generator.
So the fair statement is: the trailer motor is a range-and-comfort feature for people who plug in, not a fuel-economy feature for people who don't. An automated version that fires only during hard pulls — which is what we are proposing — would be sized to stay battery-neutral rather than to save the most fuel, and that is the design target we now have measured evidence for.
We expected the hard thermal miles to be the 100°F plains east of Denver. They weren't. The hottest the engine ever ran was on a long climb at 8,000 feet, on a day that was 79°F outside.
Sorting nearly 20,000 highway measurements by altitude, engine load, and outside temperature makes the pattern unmistakable. Climbing raises coolant temperature. Working hard raises coolant temperature. The weather forecast does almost nothing.
| What changes | Effect on engine coolant temperature |
|---|---|
| Climb 1,000 feet | +1.2 °F |
| Work the engine 10% harder | +1.9 °F |
| Outside air 10 °F warmer | no measurable effect |
Thin air is the reason. At altitude there is less air to carry heat away and less air to cool the turbochargers, and both problems arrive exactly when you're asking the engine to work hardest. The engine peaked at 241°F, about 24°F above the point where its thermostat is already wide open and the cooling system has nothing left in reserve. It spent about ten hours of the trip above that point, though only 16 minutes above 230°F and a matter of seconds at the peak — brief excursions on grades, not a sustained condition.
Why this matters to the productThe automated assist system we're proposing was originally designed to trigger on hot weather. This trip says that's the wrong signal — it would have missed the worst thermal event of the entire trip, and two of the three episodes where the engine was actively dumping extra fuel to protect itself. It needs to trigger on the climb instead. That is a design change we would not have found without driving it.
Two things we set out to measure and didn't get, stated plainly because a summary that only reports successes isn't worth much.
We wanted to split the rig's road drag into the part caused by pushing air and the part caused by tires and weight. That requires deliberately driving at 50, 55, 60, 65 and 70 mph and comparing. We used cruise control the whole trip, which meant 92% of the data sits in a single narrow speed band — not enough spread to separate the two effects, no matter how much data you collect. It's scheduled early on the next trip rather than left to chance.
The truck's data logger can see everything happening in the truck and nothing happening in the trailer. So when we say the trailer motor spends more energy than the charger returns, the direction of that conclusion is solid but the exact size rests on the trailer's published specification rather than on our own measurement. Reading the trailer's battery directly is the first item on the next trip's list.