Aston Martin is using artificial intelligence with its CoreWeave AI.R Tunnel to improve aerodynamic correlation and turn massive amounts of F1 data into faster decisions.

Aston Martin is putting AI to work in Formula 1

Artificial intelligence is becoming part of Aston Martin’s daily Formula 1 operation.

The team is working with CoreWeave to process large amounts of information and turn it into useful engineering and race-strategy insights. The project covers areas that range from aerodynamic correlation to competitor radio analysis.

The latest details were published by Aston Martin on September 21, 2026. The team says millions of data points are generated during every Grand Prix weekend, making speed of analysis increasingly important.

The objective is not to replace engineers.

Instead, Aston Martin wants its engineers to find important information faster. In Formula 1, saving even a few seconds can affect a setup change or a strategy decision.

The CoreWeave AI.R Tunnel became a central part of Aston Martin’s technical operation as the team prepared for the 2026 regulations.

The wind tunnel is at the center of the first application.

Aston Martin’s CoreWeave AI.R Tunnel provides precise aerodynamic measurements. However, a laboratory cannot reproduce every condition found at a Formula 1 circuit.

The team points to changing kerb heights, weather conditions and dirty air as examples.

Those variables can create a gap between what simulations predict and what the driver experiences on track. That difference is known as aerodynamic correlation.

AI helps Aston Martin close the correlation gap

Aston Martin describes correlation as one of the biggest challenges in modern F1 aerodynamics.

The team and CoreWeave developed analytical methods that compare historical race data with vehicle-dynamics research. The goal is to identify the factors that can cause real-world performance to differ from controlled testing.

The result is a more accurate prediction of how aerodynamic components could behave at a Grand Prix.

That information can then be used before the cars arrive at the circuit.

Aston Martin says its Vehicle Engineering team uses the insights in the week before a race to help calibrate the car’s setup. The aim is simple: make the car behave as closely as possible to the engineers’ expectations during the first practice session.

CoreWeave branding is visible on Aston Martin’s Formula 1 machinery as the partnership expands into AI-powered engineering.

The second application happens during the race itself.

Aston Martin says engineers can face 22 competitor radio channels plus more than 15 internal channels at the same time. Finding a specific piece of information manually can take valuable minutes.

That becomes a problem when a strategy decision has to be made quickly.

AI turns F1 radio into searchable data

CoreWeave has built a system that listens to, transcribes and categorizes radio communications.

The tool processes competitor and internal audio. Engineers can then search the information through a dedicated interface instead of manually reviewing long audio streams or noisy transcripts.

The system also uses a large language model to standardize terms.

For example, different ways of saying “Turn 1” can be organized as T1, making the information easier to filter.

This can help answer questions about tire performance, grip, strategy or driver feedback much faster.

Race engineers remain responsible for interpreting the data and making the decisions during a Grand Prix.

The scale of the system is significant.

At the 2026 Monaco Grand Prix, Aston Martin says the platform processed tens of hours of radio, thousands of messages and hundreds of thousands of words.

CoreWeave has also described a broader system capable of processing 40 radio channels simultaneously. Its technical report says the platform can transcribe, categorize and make those feeds searchable within seconds.

That system was tested during the final two races of the 2025 season before being deployed for 2026. CoreWeave says its development included 75 model iterations, seven hours of hand-annotated F1 radio and more than 3,000 labeled samples.

AI is another tool for Aston Martin’s engineers

Aston Martin is clear about the role of the technology.

AI is not replacing the people making decisions. Instead, it changes how much information those people can process and how quickly they can access it.

That distinction matters in Formula 1.

An AI system can organize data. It can identify patterns. It can surface relevant information.

But engineers still need to decide what that information means for the car.

The same principle applies to aerodynamic development. The AI can help predict how a component will behave, but the engineering team still decides which solution should be developed and tested.

The AI work fits a larger technical transformation

Aston Martin’s AI project is part of a much broader investment at its AMR Technology Campus in Silverstone.

The team’s new wind tunnel began operation in 2025 and was created to support both the existing car and the transition to the 2026 regulations.

Adrian Newey has also acknowledged that Aston Martin started its 2026 development cycle later than many rivals. He said the first 2026 model did not enter the AI.R Tunnel until mid-April, leaving the team roughly four months behind in that phase of aerodynamic testing.

That makes efficient use of every available development tool even more important.

The AI work is therefore not a separate experiment. It is part of Aston Martin’s effort to connect simulation, wind-tunnel data, vehicle dynamics and race information more effectively.

The Aston Martin AI and CoreWeave partnership shows how Formula 1 is moving beyond simply collecting more data. The focus is increasingly on finding the right information quickly and turning it into decisions that can improve the car on the track.

GM has officially teased its new 8.3L Duramax Turbo-Diesel V8, which will power the refreshed 2027 Chevrolet Silverado HD and GMC Sierra HD in the U.S.

GM gives its new diesel a predator-inspired name

General Motors has given its next heavy-duty diesel pickup program a dramatic first look. The company has teased the all-new 8.3L Duramax Turbo-Diesel V8, which will become the heart of the refreshed 2027 Chevrolet Silverado HD and GMC Sierra HD.

GM published the teaser on September 21, 2026, confirming the engine’s displacement and its planned application in the two heavy-duty pickups. The automaker is not releasing horsepower, torque or towing figures yet.

GM’s official teaser introduces the new “Mongoose” name for its upcoming Duramax diesel.

The new engine has been given the nickname Mongoose. GM is using the name to create a direct connection with the heavy-duty diesel competition.

The automaker specifically references a rival HD V8 diesel known as the “Scorpion.” GM’s message is clear: the new diesel is designed to compete in one of the most demanding areas of the pickup market.

For now, though, the teaser is deliberately limited. GM says this is only a first look and that Chevrolet and GMC will reveal more information soon.

That means the biggest questions remain unanswered. Buyers still need to know the engine’s horsepower, torque, towing capacity, fuel economy and availability by trim.

A new era for Silverado HD and Sierra HD

The 8.3L Duramax will arrive with refreshed versions of the 2027 Silverado HD and 2027 Sierra HD. That places the new diesel at the center of GM’s heavy-duty truck strategy for the U.S. market.

The Silverado HD is one of the two GM heavy-duty pickups set to receive the new diesel.

The Silverado HD and Sierra HD are built for customers who need serious towing and hauling capability. Their diesel engines are especially important for buyers who regularly pull large trailers or carry heavy loads.

GM has not yet explained how the 8.3-liter engine will affect the trucks’ existing chassis and towing systems. However, the larger displacement points to a major change under the hood.

The current Chevrolet and GMC heavy-duty lineup uses a 6.6-liter Duramax V8 diesel. GMC’s current HD diesel is rated at 470 horsepower and 975 lb-ft of torque, while the 2027 engine shown in today’s teaser will use a significantly larger 8.3-liter displacement.

It is important to separate those figures from the new engine. GM has not yet published official output numbers for the 8.3L Mongoose.

That makes any current horsepower or torque estimates speculative. The confirmed information is limited to the engine’s displacement, turbo-diesel V8 layout and application in the refreshed 2027 HD pickups.

The Mongoose could reshape the diesel pickup battle

The new diesel arrives during an especially competitive period for American heavy-duty trucks. Ford, Chevrolet, GMC and Ram continue to compete heavily on towing, payload, torque and long-distance capability.

GM’s decision to move to an 8.3-liter diesel V8 suggests that conventional diesel power remains an important part of its strategy for customers who prioritize maximum truck capability.

The GMC Sierra HD will also receive the new Duramax 8.3L Turbo-Diesel V8.

The Sierra HD will share the new powertrain with the Silverado HD. That means the Mongoose will serve both Chevrolet and GMC buyers, while each brand continues to give its trucks a different design and market identity.

There is also an important timing element. GM has already revealed extensive updates for its 2027 light-duty Silverado 1500 and Sierra 1500 range, including new gasoline V8 engines and an updated 3.0-liter Duramax diesel. The 8.3L Mongoose is a separate development for the heavy-duty trucks.

GM is therefore expanding its powertrain strategy across multiple truck segments. The company is retaining diesel technology while introducing new engines across its gasoline and diesel lineups.

For now, the 8.3L Duramax Mongoose remains partially hidden. GM has confirmed the engine and the 2027 Silverado HD and Sierra HD applications, but the most important performance specifications are still to come.

The 2027 Chevrolet Silverado HD and GMC Sierra HD will bring GM’s new 8.3L Duramax Turbo-Diesel V8 to the U.S. heavy-duty pickup market, marking a major new chapter for the company’s diesel strategy.

Tesla’s Cybercab is already facing navigation complaints in Austin, where one reported 10-minute trip turned into a 70-minute ride because of route restrictions and detours.

Tesla Cybercab faces a navigation problem in Austin

The Tesla Cybercab is now operating as part of Tesla’s Robotaxi service in Austin, Texas. However, some early riders are reporting an issue that Tesla owners already know well: unexpected navigation choices.

Tesla’s autonomous service is still limited in Austin. The company says Cybercab rides are currently available in limited areas of the city, while Robotaxi services also operate in several other U.S. markets.

The latest complaint involves a rider who expected a trip of roughly 10 minutes. Instead, the Cybercab reportedly took a long route through Austin and turned the journey into a 70-minute ride.

The Tesla Cybercab is currently operating in limited areas of Austin as part of the Robotaxi service.

The trip reportedly began near 99 Ranch Market and was heading toward The Domain. Instead of taking the most direct route, the Cybercab traveled south and continued through surface streets.

That difference can be significant in Austin. The city relies heavily on major highways for longer trips. A route that avoids those roads can become much longer.

Current reports indicate that Tesla’s Robotaxi operation has been avoiding highway travel during its early Austin deployment. The company, however, describes Cybercab as a vehicle designed to handle city streets, highways and complex intersections.

This creates an important distinction. The vehicle may be technically designed for highways, but the current operating area and operational restrictions can still limit which roads it actually uses.

Railroad crossings could make routes even longer

Highways may not be the only factor. Austin also has several railroad crossings that can complicate autonomous routing.

Some riders have suggested that avoiding certain at-grade railroad crossings could force Robotaxis to travel farther before finding a suitable crossing point. That possibility has not been confirmed by Tesla.

Still, the combination of road restrictions and railway crossings could explain why a relatively simple trip became so complicated.

The rider also reported that the Robotaxi app did not provide a clear estimated travel time before the trip began. Tesla’s current service is therefore placing extra importance on how its vehicles calculate routes once the ride has started.

The driverless Cybercab operates without a conventional steering wheel or pedals.

There have been other reports of unusual routing. Another passenger reportedly found that the Cybercab stopped at a business next to the requested destination rather than directly at the intended location.

That does not necessarily mean the same problem affects every ride. There is still limited public data about how frequently these incidents occur across the entire Robotaxi fleet.

However, the reports show how different autonomous navigation can be from conventional navigation.

A human driver can recognize that a route looks inefficient. They can also make a quick change, turn around or choose a different road. A fully autonomous vehicle must operate within the rules and boundaries programmed into its service.

Tesla’s navigation history adds another layer

Navigation complaints are not entirely new for Tesla. Owners have discussed routing errors and unexpected directions for years.

In a conventional Tesla, those mistakes can be frustrating but relatively easy to correct. The driver can simply ignore the route and choose another road.

That option changes when there is no driver behind the wheel.

The Cybercab is designed as a fully autonomous vehicle. Tesla’s official documentation describes it as a two-seat vehicle without a steering wheel or pedals, using cameras and sensors to navigate its environment.

That makes navigation accuracy much more important.

If a route is inefficient, the passenger cannot simply take control. Instead, the rider has to remain inside the vehicle or end the ride and find another way to reach the destination.

The Cybercab is Tesla’s purpose-built autonomous taxi, designed for fully driverless operation.

Tesla is also taking a gradual approach to its Robotaxi expansion. The company began offering employee rides before launching wider autonomous service and has continued expanding its operating areas step by step.

The limited operating environment is important because autonomous vehicles normally begin with a defined operational design domain. That can include restrictions involving geography, road types, traffic conditions and other factors.

As those boundaries expand, some of today’s routing problems could become less common.

For now, though, Austin provides an early look at one of the practical challenges facing Tesla’s autonomous strategy. The issue is not necessarily the Cybercab’s ability to move through traffic. Instead, it can be the route chosen before and during the journey.

Tesla’s Robotaxi service is already operating in the U.S. market, with Austin serving as one of its key early locations. The Cybercab itself is currently available only in limited areas of Austin, making the city an important testing ground for Tesla’s driverless transportation plans.

For passengers, the biggest question is simple: how quickly can Tesla improve route selection while keeping its strict autonomous operating limits? The answer could have a major impact on how useful the Cybercab becomes for everyday trips.