Cummins Confidential : The Triple Drop – Fix The Derate, Flog The Diesel, Feed The AI

By /

Cummins laid out three courses of corporate optimism on 15 October 2026, starting with a diesel engine learning not to punish its owner quite so savagely. Next came a 10-litre bus engine with another engagement on the European publicity circuit. Dessert arrived inside an artificial intelligence data centre. There, the supposedly weightless digital future needs an extraordinary amount of heavy fucking machinery to keep breathing.

The menu promised customer support, engineering progress and dependable power; the accompanying paperwork offered a considerably less wholesome tasting menu. A service notice contained a warranty exclusion, the bus had appeared at an exhibition a year earlier, and an investor presentation exposed a multibillion-dollar appetite for generators. Presumably somebody in corporate communications had paired the meal with a bottle of Destination Zero.

This is the pleasure of reading a corporate newsroom alongside its less glamorous paperwork. Customers receive comforting adjectives, engineers get fault codes and investors learn what the equipment might earn over its working life. The important details are in the company’s own handwriting, available to anybody willing to look beyond the polished fucking headlines.

DEF Inducements: Five Miles Per Hour Through The American Dream

Picture an American truck driver after ten hours on the road. The coffee tastes like someone rinsed a crankcase into the cup. Freight needs delivering, the engine still has power and the schedule is already in trouble when an emissions-system fault appears on the dashboard. Under certain previous inducement conditions, the truck could eventually face a five-mile-per-hour speed limit.

Five miles per hour. You could stop to examine a particularly interesting pothole and scarcely lose momentum. The expensive commercial vehicle has become a disciplinary exercise with mirrors, while dashboard software decides that earning a living must proceed at walking pace. What a magnificent, fucking stupid way to spend an afternoon.

Selective catalytic reduction systems use diesel exhaust fluid to reduce nitrogen oxide emissions. Operators must maintain that equipment, and regulators need effective enforcement. Fine. But compelling repairs and rendering a working truck virtually useless are not quite the same proposition. The delivery deadline doesn’t surrender merely because the engine computer has developed a stern personality.

The Great Diesel Pardon

Cummins’ first release bears the benevolent title “Cummins Continues to Support Customers with Updated DEF Inducement Calibrations”. Following EPA guidance, certain updated calibrations give operators longer to resolve faults before restrictions become severe. The final speed limit rises from five to 25 mph for applicable trucks, while eligible motorcoaches can reach 50 mph.

These are useful changes: a truck moving at 25 mph has a much better chance of reaching help without a recovery vehicle. But Cummins is implementing regulatory flexibility, not handing down a personal act of mercy. EPA issued the guidance in August 2025; Cummins announced its programme in July 2026, then returned for another reassuring October update.

Brett Merritt, president of the Engine Business, speaks about helping customers get their jobs done. An admirable ambition, although the people paying for those engines might ask what the old restrictions cost them. Lost hours, missed deliveries and recovery bills do not evaporate when the newsroom publishes a warmer explanation. The update deserves attention; so does the damage the earlier arrangement could inflict.

The Accountant Enters The Workshop

Cummins’ QuickServe notice has rather less interest in emotional reassurance. It deals in engine families, model years, electronic control modules and release dates. X12, X15, L9 and B6.7 engines occupy a schedule stretching into December 2026, precisely the detail technicians need when an actual machine arrives at the workshop. They do not need a photograph of somebody smiling beside a truck.

Among the calibration details sits an exceptionally helpful sentence: “This change is not covered under warranty”. Ah, the accountant has entered the workshop, removed his coat and settled beside the invoice printer. That notice does not establish that customers must pay. Cummins might fund the work through another arrangement, but the public announcement never explains who covers the calibration, labour or any related diagnostics.

If Cummins pays, why not say so? And if the operator pays, what exactly does this expression of customer support cost? Instead, the company sends owners towards their local dealer or distributor. Columbus supplies the reassurance. Someone behind a service counter gets the awkward conversation, while the customer waits to discover whether the word “support” comes with an hourly rate.

A Programme Without A Proper Scoreboard

Cummins said in July that it had begun offering updated calibrations for more than 1.5 million engines. The October release reports that tens of thousands have received them. Those are different measures – an eligible population and completed installations – so they cannot produce a reliable uptake percentage. But they certainly justify asking for one, along with a timetable broken down by engine family.

The company monitors minute details of combustion, exhaust treatment and electronic control. Yet its customer-facing progress report offers an impressively elastic quantity. How many owners has Cummins contacted, and which releases remain outstanding? What will still be waiting in December? Surely somebody can assemble a fucking table without asking the diesel industry to invent mathematics again.

Meanwhile, EPA’s July 2026 proposal goes further than these calibrations. For newly manufactured equipment, it proposes warnings in place of mandatory SCR-related derates and speed restrictions. The agency is also considering a route for older engines, although this remains a proposal rather than permission to ignore emissions requirements. Still, the regulator is questioning a punishment mechanism that once reduced a commercial vehicle to pedestrian speed. Cummins can celebrate the improvement. Customers deserve the bill, the timetable and the reason it took so long.

X10: Another Lap Around The Exhibition Carpet

Now leave the truck stop for a European motor show. The carpet has that strange conference-centre smell, somewhere between adhesive and boiled coffee, while engines sit beneath heroic lighting. Sales representatives polish their smiles while visitors inspect metal parts that supposedly change civilisation. Every stand has a future to sell, preferably with a glossy brochure and complimentary pen.

Cummins’ second release announces “New X10 moves the bus game forward”. Its opening remembers visitors encountering the X10 Euro VI at Busworld Brussels. That event took place in October 2025. Cummins displayed the engine on Stand 810 in Hall 8, where journalists duly covered its launch and specifications. Now the October 2026 article revisits an introduction from a year ago.

There is nothing wrong with revisiting an engine as development continues. There is something funny about a newsroom treating the familiar as if it has just emerged from the packaging. The X10 has travelled from Brussels through another year of publicity, only to arrive back at the same departure gate. Somebody should check whether its exhibition badge now qualifies for fucking loyalty points.

Actual Engineering, Questionable Calendar Management

The machine itself deserves better than a lazy dismissal. Cummins quotes 355 to 454 horsepower for the Euro VI X10, with torque reaching 2,200 Nm. Its published claims include a six-to-eight-per-cent fuel-economy improvement against current comparable engines. Reports from Busworld also put its weight saving at around 150 kilograms. Coach operators have good reasons to care about all four figures, especially when a vehicle spends years climbing hills with passengers and luggage.

Yet a claimed saving needs a usable comparison. Which engines provided the baseline? Did the tests resemble loaded coaches climbing long gradients in bad weather? What remains of the advertised improvement after years of ordinary service? A fleet manager spending serious money needs more than the manufacturer’s confidence and a nice photograph under the exhibition lights.

By June 2026, Cummins was already promoting an X10 Euro 7 at Busworld Türkiye. Its material quoted 320 to 450 horsepower and up to 2,299 Nm of torque. The platform has genuinely developed, but this public timeline makes October feel less like a discovery than another clean plate for yesterday’s lunch.

Which Bus Game, Exactly?

There is a more important distinction beneath the showroom comedy. Euro 7 concerns regulated emissions, while diesel combustion still produces carbon dioxide. Cleaner is useful; zero-emission is something else, especially when Cummins discusses bus transport as though every route uses the same technology.

Transport & Environment says zero-emission models represented 60 per cent of new EU city-bus sales in 2025. Battery-electric models accounted for 56 per cent, with fuel-cell vehicles taking another four. EU rules require 90 per cent of new city buses to be zero-emission by 2030. The requirement reaches 100 per cent in 2035. Urban bus operators are already planning around chargers, batteries and depot electricity.

The X10 mainly targets interurban buses and coaches, which face different rules and operating demands. An efficient diesel engine can still make commercial sense there; its continued existence is not the scandal. The question is what Cummins has actually advanced, for which customers, and how its performance claims survive outside the brochure. A good coach engine deserves a measured case. It doesn’t need another fucking coronation beneath the fluorescent lights.

AI Data Centres: The Transistor Misses Its Own Birthday

Cummins’ third article, “From strategy to standby”, begins with a lesson in technological history. According to its newsroom, Bell Labs invented the transistor in 1948. Unfortunately, John Bardeen and Walter Brattain demonstrated the first working point-contact transistor in December 1947; Bell Labs announced it publicly in June 1948. Cummins has confused the invention with the press conference.

That’s an exquisite opening for a company explaining the engineering behind artificial intelligence. Perhaps the transistor sat in an office for six months awaiting clearance from corporate communications. The scientists had built a working device, but apparently it didn’t count until somebody approved the fucking announcement. Scientific history has finally adopted the newsroom calendar.

From there, Cummins travels through computers, the internet and the demands of AI. It is an ambitious route from germanium to generator sales. The article eventually reaches its real subject: data centres need reliable power. The digital age may sound weightless, but its machines sit in buildings full of cables, heat, switchgear and mechanical contingencies.

Underneath The Cloud, A Sixteen-Cylinder Bastard

Cummins offers Centum Series generators and the QSK95, built around a 95.3-litre V16 engine. Forget the ethereal pictures of artificial intelligence floating somewhere above the weather. Here is a great collection of pistons, crankshafts, coolant and electrical equipment. Its job is to feed vast banks of processors when their normal electricity supply fails.

The contrast is magnificent. Humanity has built systems capable of writing poetry, analysing scientific papers and producing an industrial quantity of digital bollocks. Yet somebody still needs to check the fuel, plan the maintenance and keep a technician within reach. The computer can discuss consciousness while the engine demands lubrication, fuel and a starter that works.

Brandon Ewing, Cummins’ data-centre strategist, wants the company involved years before a facility opens. It can help developers choose generators, batteries, cooling systems and supporting infrastructure. Cummins also offers parts, servicing and agreements covering equipment from other manufacturers. That makes sense for customers who cannot tolerate outages, but an early meeting can also lead to years of commercial involvement. The hardware arrives once. The relationship can keep earning long after the delivery lorry leaves.

The Investor Menu Has Better Numbers

On 21 May 2026, Cummins told investors what this opportunity might be worth. Management expected around $5 billion in data-centre revenue in 2026. Its target exceeded $9 billion a year by 2030, across Power Systems and Distribution. That is revenue rather than profit, and the later figure remains a target rather than cash already collected. Even so, the ambition makes the October reassurance look positively modest.

Cummins also announced another $450 million for manufacturing expansion. It plans to add 20 gigawatts of high-horsepower engine and generator capacity, reaching 55 gigawatts by 2030. These figures concern manufacturing, not electricity already flowing through operational data centres; the scale remains extraordinary. Someone is preparing a very large factory response to an industry famous for describing itself in weightless, digital terms.

The investor story has no need for sentimental descriptions of uninterrupted internet access. It discusses production capacity, distribution, customer demand and future service income. Engineers must keep the systems running, while accountants calculate what that will earn. At least the investor presentation lets them speak without wrapping every sentence in a bloody sustainability waistcoat.

It Rarely Runs. Until It Does.

Power Systems president Jennifer Bush explained that around 95 per cent of Cummins’ generation business involved standby applications. She expected standby to remain dominant through 2030. A backup generator usually waits, undergoes occasional testing and starts when the normal supply fails. In Bush’s words, “it rarely runs”. The owner buys reassurance in the shape of expensive hardware that everyone hopes never to need.

Prime power changes the relationship. Cummins is pursuing applications where generating equipment runs far more regularly, sometimes alongside constrained electricity grids. More operating hours mean fuel consumption, component wear and servicing; Bush compared the opportunity with mining engines rebuilt several times during their working lives. The opportunity does not end when Cummins sells the generator. It extends through the years of keeping that generator alive.

Standby still dominates the business, and Bush expects that pattern to continue through 2030. The additional prime-power market offers something different: heavier duty cycles and greater aftermarket demands. Once generators work rather than wait, the parts department has more reasons to stay busy.

The Calendar Develops An Aftermarket Opportunity

Then came the detail that deserved a full stop and a long look. Bush suggested a heavily utilised prime-power engine might run for 10,000 hours a year. Unfortunately, the year offers only 8,760. The published transcript contains an extra 1,240 hours that the planet has neglected to supply.

Perhaps she was speaking loosely, or the transcript missed a correction. Either way, Cummins has accidentally proposed an engine so busy that it needs an aftermarket supplier for the fucking calendar. The arithmetic is absurd, but the commercial point underneath remains clear. Higher utilisation means more servicing and potentially more rebuilds. The company has already explained why it wants to be there for every one of them.

The number is a spectacular cock-up, but the sales logic is perfectly coherent. As running hours accumulate, the owner buys maintenance and the manufacturer supplies it. The data centre needs dependable electricity; Cummins can invoice for the parts and labour that keep it flowing.

Texas: The Cloud Meets The Gas Field

The strategy already has a concrete example. In June 2026, Circe Energy announced an order for about 2 gigawatts of Cummins natural-gas generating capacity. Deliveries are scheduled between 2026 and 2030. The equipment will support its planned West Texas AI campus and possible future developments. It is an order, not 2 gigawatts already operating, but hardly a speculative doodle on a conference slide.

West Texas offers the kind of geography that strips the glamour from digital utopianism. There is land, gas and the prospect of dedicated generation beyond a conventional grid connection. Fossil fuel can travel into a generator and emerge as electricity for some of the world’s most advanced computing equipment. An industry obsessed with tomorrow has found an old accomplice beneath the Permian Basin.

Cummins also sells battery storage and integrated electrical systems. Its August 2026 data-centre battery announcement demonstrates that broader portfolio. Good. Those technologies deserve comparison on actual use, not tribal loyalties to a favourite machine. A generator waiting for an outage has different emissions from a gas plant supplying routine power. The relevant figures are operating hours, fuel, efficiency and permitted emissions. The cloud doesn’t float. It squats on concrete, feeds on megawatts and leaves an environmental bill that won’t disappear into thin fucking air.

Destination Zero Meets The Order Book

Cummins has never promised to stop selling combustion engines overnight; Destination Zero explicitly allows a phased transition and improvements to existing technology. There is little value in attacking a promise the company never made. The stronger question comes from its own expansion plans: how will this much additional machinery affect lifetime emissions?

A company can make an individual engine cleaner while selling more engines overall. It can install batteries beside generators while also developing new gas-fired prime power. Neither fact settles the environmental arithmetic, although Cummins already measures financial opportunities with impressive care. It should apply comparable precision to fuel consumption, equipment usage and progress towards its climate commitments.

That is where the three announcements finally meet. Software extends the usefulness of existing diesel equipment. A newer engine serves the coach market. Industrial generation opens another huge source of revenue and service work. Each decision has a customer rationale, but together they show a business expanding its combustion interests while promising a lower-emissions future. The accounts show where the orders are heading; Cummins should be able to show whether its environmental targets follow.

Last Orders From The Newsroom

Cummins gives three different customers what they want to hear. The truck operator hears that work need not stop quite so suddenly. The coach manufacturer gets another promise of performance, while the data-centre developer hears that power will be waiting. These are real problems with real machinery behind them. The company’s investors, however, hear a different melody: equipment sales, additional capacity and years of servicing.

The unexpected common ingredient is time. A derated truck costs its operator hours that nobody refunds. Last year’s bus launch acquires another afternoon beneath the newsroom spotlight. At the data centre, more running hours can mean more maintenance income for the equipment supplier. Time is the one thing every customer wants to save, and the one thing Cummins can keep finding ways to sell.

Its figures are precise when money enters the room. Try applying the same discipline to update costs, installed calibrations, real-world fuel savings and lifetime emissions. The answers might make a finer story than another glossy procession of engineering promises. Until then, the newsroom can keep laying the table. Somewhere behind it, the invoice printer has already warmed its fucking rollers.

Fix the derate. Flog the diesel. Feed the AI. Then send the invoice.

Lee Thompson – Founder, The Cummins Accountability Project

Sources

Scroll to Top