A 90% Fuel Saving That Weighs Ten Tonnes

There is a piece of hardware in the United States that turns an ordinary diesel tractor-trailer into a hybrid without touching the engine.

It is a dolly. It couples between the tractor's fifth wheel and the cargo trailer, and it contains a battery, an electric motor and a driven axle. The rig hitches up in about the time it takes to couple any trailer, and the driver does not have to do anything differently afterwards. When the pack runs down, the whole unit is swapped at a roadside hub in under five minutes.

The company says it cuts diesel consumption by seventy to ninety per cent, taking a truck from six to eight miles per gallon to around twenty.

I want to take that claim seriously rather than dismiss it, because the underlying engineering is genuinely clever and the verified numbers are good. The problem is not that the claim is false. It is that miles per gallon cannot answer the question a fleet is actually asking — and the test house says so itself, in the certification.

What Independent Testing Actually Measured

The unit has been through certified third-party testing, which is more than most claims in this category can say.

On a 41-mile loop at 65 mph, with a sleeper cab tractor and a 53-foot dry van at roughly 79,000 lb gross, the measured improvement in fuel economy was 65.2%, at 14.79 MPG. Across the range of driving scenarios tested, the improvement fell between 65.2% and 69.7%, or 14.79 to 19.12 MPG.

So the independent figure is about two-thirds, not ninety per cent. The higher numbers in the marketing are attributed to favourable routes and later software tuning, and the company is reasonably explicit that terrain determines the result.

A two-thirds reduction in diesel burn on a Class 8 truck, with no engine change and no driver retraining, is a serious number. I do not want to lose that in what follows.

The Sentence in the Certification

Buried in the test house's own write-up is the line that reframes everything above it.

The improvement in fuel economy, they note, is a tradeoff for electrical power and charging, which should be accounted for.

That is a test house telling you, in its own certification, that the number it has just certified is incomplete.

And it is obviously right. The diesel consumption fell because roughly half the propulsion energy now arrives from a socket instead of a tank. Miles per gallon measures one of the two energy inputs. It was designed for vehicles that only had one.

The size of the missing input is not hidden. In the same testing, the unit covered 123.3 miles while the state of charge fell from 98.8% to 40.0% of a 575 kWh pack. That is a little over 2.7 kilowatt-hours of electricity per mile — energy that was generated somewhere, paid for by someone, and appears nowhere in a figure of 14.79 MPG.

I deliberately will not turn that into a headline efficiency percentage, because the honest position is that the published data does not support one. The fuel economy and the battery consumption come from different test segments, and the answer moves depending on which energy content you assume for diesel. That is not a gap in the reporting. It is the finding: the metric everybody quotes is structurally incapable of telling you whether total energy went up or down.

Which matters, because total energy is what a fleet pays for and what a regulator eventually counts.

Ten Tonnes Has To Come From Somewhere

The second thing the headline does not carry is mass.

The dolly is about four metres long and weighs roughly ten tonnes, most of it a 575 kWh lithium iron phosphate pack. IEEE Spectrum's write-up puts the consequence plainly: the ten-tonne dolly eats into payload capacity — a nonstarter for weight-sensitive loads.

That single constraint sorts the entire addressable market in two.

For freight that fills the trailer before it reaches the weight limit — parcels, packaging, most light consumer goods — the mass is close to free. You were never going to use that capacity.

For freight that hits the weight limit first — liquids, steel, cement, aggregates, most bulk — every tonne of dolly is a tonne of revenue that does not travel. You do not need a spreadsheet to see how that ends.

Note the test conditions again: roughly 79,000 lb gross, against a US federal limit of 80,000 lb on the interstate network. The test was run essentially at the ceiling. That is not a criticism of the test; it is a demonstration of where the constraint bites.

I find this the most useful part of the story, because it is the part that transfers. Almost every efficiency technology in commercial vehicles arrives with a mass penalty, and the question of whether it is affordable is never a question about the technology. It is a question about what the customer hauls.

The Part That Is Hardest to Engineer

There is a third constraint that has nothing to do with the vehicle.

The concept only works with a network. A pack that gives a couple of hundred miles of assistance needs swap hubs positioned along the route, and a swap hub is only viable if enough trucks pass it. That is the ordinary chicken-and-egg problem of any refuelling infrastructure, and it is usually harder than the engineering.

It also explains something about the product roadmap that would otherwise look like a retreat. The next generation is reported to carry a smaller battery while holding the same usable range. Read alongside the payload problem, that is not a downgrade. It is the correct direction — and it suggests the company has worked out which constraint is actually binding.

What I Would Ask Before Buying One

None of this makes the technology a bad idea. It makes it a specific idea, which is a different thing, and far more useful to know.

If I were assessing it, I would not start by challenging the 65%. I would ask four questions the headline cannot answer.

What is the total energy per tonne-kilometre, diesel and electricity together, on my routes? Not MPG. If a supplier cannot produce that number, it is because the figure they lead with was never designed to contain it.

Is my freight weight-limited or cube-limited? This one question decides whether ten tonnes is nearly free or disqualifying, and it is answerable from existing load data this afternoon.

What does the electricity cost where the swap hubs are? Diesel displacement is only a saving at the delivered price of the thing replacing it — and industrial tariffs vary more between regions than diesel does.

What happens on the day the hub is down? A hybrid that degrades to an ordinary diesel truck carrying ten tonnes of dead battery is a very different risk from one that simply stops.

For Indian operations there is a fifth, and it is the one I would ask first. Our gross weight regimes, road conditions and route structures are not the American interstate network, and a concept whose entire economics rest on swap-hub density has to clear a much higher bar in a market where that density does not exist yet.

The Reason I Wrote This Up

The lesson here is not about battery dollies. It is about which number a technology chooses to lead with.

Miles per gallon is not a dishonest metric. It is a legacy one. It encodes an assumption — that a vehicle has exactly one energy input — which stopped being true the moment anything was electrified. Every hybrid architecture flatters itself on that metric automatically, without anyone intending to mislead, because the metric was built for a world with one tank in it.

That is worth remembering the next time a supplier walks in with a percentage. The useful question is rarely whether the number is true. It is what the number was built to measure, and whether that is still the thing you need to know.