Autonomous Vehicles

Automated driving is back on the industry’s front burner, a resurgence led primarily by advances in artificial intelligence and marketplace success, industry insiders said at CES 2026 in Las Vegas in early January.

Automaker interest in pursuing top-tier Level 4 and 5 automated driving cooled in recent years as the technology hit some speed bumps, OEMs backed away from mobility-service schemes and available product-development dollars largely were directed toward battery-electric vehicles instead.

But with a policy-driven slowdown in the U.S., continued cost disadvantages compared with internal-combustion-engine vehicles and a consumer base still reluctant to make the switch to full-electric vehicles in big enough numbers, industry investment has begun to flow away from BEVs and back toward automated driver-assistance system technology.

But it isn’t just the lower boil in BEV activity that is sparking renewed ADAS interest, executives told WardsAuto.

“I think automakers worldwide are seeing the benefit automated driving can bring,” said Rajat Sagar, vice president-product management for the Snapdragon Ride Pilot platform (developed with BMW) at chip and software-solutions provider Qualcomm.

Automakers in China and others such as Tesla have been deploying ever more-sophisticated ADAS capabilities as a way to entice consumers, Sagar noted. “That’s what is fueling the need for these implementations.”

Why automakers are investing more in ADAS

C.J. Finn, U.S. automotive leader for consultant PricewaterhouseCoopers, contends automakers now see being among the first to offer higher levels of automated driving as the best near-term opportunity to gain market share.

“There’s a constant fight of what’s that [next] technology where if you’re the first mover you can really capture market share,” Finn told WardsAuto. “Some thought electrification may have been it, but autonomous is that next thing ... When you move from Level 2-plus to Level 4, that’s the game changer.”

Vlad Voroninski, CEO at California startup Helm.ai, agrees the money is once again flowing into automated driving.

“OEMs [realize] the main revenue driver is going to be in assisted driving for commercial vehicles. That’s where the capital's flowing,” he said. “It is probably the No.1 topic in the boardrooms of automotive companies.”

Even in the U.S. market, there’s evidence buyers are beginning to see a real demand for the technology from consumers. Researcher AutoPacific said in July that 43% of U.S. car buyers point to hands-off semi-autonomous driving for highway use — Level 2-plus systems such as General Motors’ Super Cruise, Ford’s BlueCruise and Tesla’s Full Self Driving — as the technology they want most in their next vehicle. That marked a hefty 20-point increase from a year earlier.

 A 2023 McKinsey study showed two-thirds of consumers would be willing to pay a $10,000 premium for Level 4 automated driving technology (hands off and eyes off on both highway and urban roads) on their next vehicle. Precedence Research predicts the autonomous-vehicle market could top $2 trillion annually by 2035, growing more than 35% per year between now and then.

How GenAI is helping build the momentum

Making all this possible is the rapid advancement in technology — thanks in big part to the leap in AI capability that has altered the approach to self-driving software development, and the improvements in — and proliferation of — sensor technology that is bolstering the quality and volume of data required to perfect ADAS operation. Mobileye, for example, said it collects data in real time from some 8 million vehicles globally, harvesting 32 billion miles (51 billion km) of data in 2025.

“What we’re doing today isn’t something we just decided to do yesterday,” said Manuela Locarno Ajayi, a product-tech executive with TomTom, noting with half the cars on the road now connected, there’s a wealth of real-world performance data available for developers to help fine-tune their software. “We’ve been trying to do it for a long time. Only now the technology is catching up to be scalable and cost-effective [and] to have this level of precision.”

TomTom, which used CES 2026 to show off its next-generation high-definition mapping for navigation systems it is supplying to Volkswagen’s CARIAD software-development group, among others, said it now can map 470,000 miles (750,000 km) per day thanks to the advances in AI technology and the growing amount of real-world data available.

“I think the technology is closer to being ready,” Jagan Rajagopalan, head of strategy and portfolio for integration and software specialist Elektrobit, said in explaining the recent surge in ADAS interest from automakers. “It’s just AI being ready; the lidar, the cameras and technology getting more standardized. In 2030, you’ll begin to see more mainstream autonomous vehicles.”

The arrival of Generative AI and development of visual language models has changed the way automated-driving software is being written and tested, industry insiders said. With VLMs, AI software now is capable of understanding and acting on image inputs, rather than just language.

“Overnight we can train [virtually] with data that no real-world data can match,” said Mobileye CEO Amnon Shashua, adding Mobileye can run billions of simulations a day.

“The models you’re seeing now are designed to be end-to-end models,” said Nakul Duggal, executive vice president and general manager of automotive for Qualcomm, meaning the software thinks more like a human driver and can analyze the scene and determine the vehicle’s path much more quickly now. “That evolution in AI has only happened in the last three to three-and-a-half years. The data has existed but the visual-language models didn’t.”

 “As you train those newer models, you get much more robust outcomes,” he said. “Once you have the recipe in place, which is not easy to get to, then its [simply] a data-driven obstacle.”

In earlier attempts at solving autonomous driving, Helm.ai’s Voroninski said: “You would have a very serious ceiling on what you could do. You might never be able to offer Level 3 or Level 4 features with that approach.”

Now, with its perception stack said to function more like a human brain, there’s line of sight to the ultimate goal.

“Seventy percent of your brain is fully dedicated just to being able to perceive the world,” Voroninski said. “That’s a problem we’ve spent years solving and have effectively solved from an automotive production-grade perspective.

“We've shown that we can actually train an autonomous driving system that can do 20-minute drives in cities it's never seen before, because we have this production-grade perception stack.”

NVIDIA President CEO Jensen Huang summed it up succinctly. “You no longer program the software, you train the software,” he said, pointing to his company’s new Alpamayo open-source platform launched at CES 2026 that underpins the automated drive stack deployed on the Mercedes CLA. He says $100 trillion of global R&D has shifted to AI from classical methods of product development in the last few years.

As costs fall, ADAS could be built for the masses

While the technology is improving, costs are declining, meaning higher-end Level 2-plus capability will begin to filter down to more mass-market vehicles. Ford’s move to develop its own technology in-house is just one example of the drive to get Level 2-plus functionality into entry-level vehicles.

Mobileye, for instance, said the program for its Surround ADAS platform with a U.S. automaker beginning mid-2028 will cover up to 9 million vehicles. That comes on top of a similar program Volkswagen announced in 2025 that is expected to include up to 10 million Surround ADAS-equipped vehicles.

“This is a very cost-efficient solution,” said Nimrod Nehushtan, executive vice president-business development and strategy for Mobileye. “These are two of the top six automakers in the world that are sourcing their high-volume models with this solution. It shows there’s a trend, that there’s a lot of focus on the cutting-edge performance in the premium segments, but there is also an interesting development coming from high-volume cars.”

Qualcomm says its Ride Flex is another less costly platform gaining traction. It can combine cockpit infotainment functions and ADAS controls on a single system-on-chip, making it well-suited for lower-priced vehicles. Its first production application debuted in 2025 on several Chinese models, including BAIC’s ARCFOX Alpha T5, General Motors’ Buick Electra L7 and Dongfeng Nissan’s N6, and Qualcomm says it has more than 10 partners with programs in the works utilizing the Ride Flex SoC, including a collaboration with Hyundai Mobis.

“What this means is, you are starting to see the implementation of ADAS and [in-vehicle infotainment] across almost every tier of vehicle,” Qualcomm’s Duggal said. No vehicle segment is being left out of the movement for Level 2-plus and above autonomy, “which increases the market size for us and everybody else,” he added.

Costs are certain to come down further as volumes rise, executives contend.

Shashua said Mobileye is targeting a 40% cost reduction for its L2-plus Surround ADAS system in 2028 and will continue to lower costs of its entire suite of technology beyond that. With its Chauffeur Level 4 software VW is deploying in its MOIA fleet, Mobileye is looking to slash costs by reducing the number of lidars required from three today to just one.

After years of promise, some developmental stumbles and on-and-off interest and investment from global automakers, it appears automated driving is gaining a solid foothold.

“I’m very certain in the next 10 years a large percentage of cars will be autonomous, or highly autonomous,” NVIDIA’s Huang predicted.

Dive Brief:

• General Motors has started testing its next-generation, “eyes-off” autonomous driving technology on public roads with development vehicles operating on limited access highways in California and Michigan, the automaker announced in a March 23 blog post.
• GM said it will soon evaluate its self-driving tech with over 200 test vehicles in real-world driving environments. Each vehicle will have a trained test driver behind the wheel who can take manual control of the vehicle at any time.
• The automaker announced its next-generation autonomous driving technology in October 2025, which is a more advanced, Level 3 version of its Super Cruise hands-free highway driving system. The upgraded system will debut in 2028 in the Cadillac Escalade IQ electric SUV before rolling out to more GM vehicles.

Dive Insight:

GM said in its blog post that the launch of supervised on-road testing “marks a significant transition from manual data collection,” since it includes activating its eyes-off driving system in real-world driving environments to evaluate its capabilities. The automaker says that on-road testing “is essential to build a trustworthy system.” 

GM will launch the self-driving system for highway driving initially, before offering “driveway-to-driveway” capabilities. In March 2025, GM said it was using Nvidia’s Drive AGX platform for the development of its next-generation autonomous driving technology.

In its future models, GM aims to support Level 3 automated driving, which provides hands-off, eyes-off driving but requires the driver to be ready to take over. To support this, it’s adopting a new centralized computing platform starting in 2028. The software-defined-vehicle architecture bundles all of a vehicle’s core systems, including powertrain, steering, infotainment and safety, using a single, high-speed processor. It will allow GM to more easily add eyes-off driving capabilities to other models without having to rebuild it from scratch for each respective application, the automaker said in its blog post.

The new SDV platform adds 10 times more over-the-air software update capacity to GM vehicles, as well as 35 times more AI performance to support autonomous driving and other advanced features, according to the company.

GM’s future centralized compute platform will also be continuously improved over time via over-the-air software updates from data collected from its entire fleet. This data is continuously being fed directly back into its autonomous driving development cycle to improve AI driving models and overall system robustness. 

For example, construction zones or degraded lane markings flagged by its test vehicles are added as data points that feed back into the autonomy development loop for the entire fleet, continuously improving the system’s performance over time, GM said in a separate engineering blog post March 19.

To date, the automaker said it has accumulated more than 800 million customer-driven miles logged across 23 vehicles equipped with Super Cruise, 5 million of which were fully autonomous miles in what the company called “some of the most complex urban environments in the country.” 

GM then combines the vehicle data it collects with testing in computer simulation, which the automaker says is equivalent to approximately 100 years of human driving experience “every single day.” The company said it's focused on accelerating the learning rate of its autonomous driving system, which also boosts safety.

In addition to launching its new eyes-off autonomous driving system, GM plans to add conversational AI to its vehicles with the integration of Google Gemini. Drivers will be able to use natural conversational language to interact with their vehicles. GM also plans to introduce a proprietary version of generative AI that will draw from vehicle-specific information, the company said in 2025.

To prepare for the rollout of more advanced autonomous driving, GM reshuffled its software and services unit in November, tapping Sterling Anderson, a co-founder of autonomous trucking developer Aurora Innovation, to lead the reorganized unit. Anderson previously served as chief product officer for Tesla, where he led the team that delivered the company’s Autopilot autonomous driving system.

Sponsored content

By TaskUs

Successful expansion into Japan depended on building a team that understood both technology and local nuance. 

Japan has one of the world’s most complex road environments. Historically, the confusing and fragmented network of streets served as a territorial defense. The country’s topography, space constraints and strict local road rules make it one of the most difficult places to scale an autonomous fleet. 

For a leader in embodied AI, breaking into this market called for operational precision that met both internal safety goals and strict local regulations. Success in autonomous vehicle (AV) commercial deployment depended on building a team that understood both the technology and local nuance — a savvy group of bilingual experts who could sit behind the wheel and provide real-time data inputs and interventions to improve the software and safety features.  

The challenge: A model that couldn’t scale 

The company’s initial attempt to scale in Japan revealed a gap between what the market demanded and what its existing partner could deliver. The former partner operated more like a staffing agency, sourcing people but not providing capabilities.  

Having to manage coaching and team development pulled the focus away from the company’s core mission of advancing the technology. And given how specialized these roles were, that was a significant burden.  

Without a partner who could own the entire process, building a team capable of meeting the demands of the region, as the speed of growth, wouldn’t be possible. 

The solution: A fully managed AV program 

From driver ops to depot management, the company turned to TaskUs to take control of the process. Rather than adapting a generic model to fit, we built a program designed around the specific demands of Japan’s market. 

Getting the knowledge on the ground 

The foundation started before a single local hire was made. We flew in seasoned U.S.-based operators, ops leaders and trainers who had supported major AV clients. They stayed on the ground for months — long enough to transfer best practices and localize all training materials. Nothing was assumed to translate directly. 

Rethinking the hire 

With a strong foundation in place, we shifted focus to building the right team. Safety was the first criterion for hiring, and every operator candidate needed a minimum three-year clean driving record. Beyond that, each hire completed a specialized driving certification run by a local driving school, calibrated to the client’s precise safety requirements. 

Experience in the vehicle itself, however, almost became an obstacle. Our data showed that seasoned drivers — former taxi drivers, for example — often struggled with the digital tools central to the role. We adapted the profile toward tech-savvy professionals who could be developed into capable operators, rather than experienced drivers who needed to be retrained. 

For depot roles, our internal referral program helped source the bilingual mechanics and technicians the client needed — specialists not easily found through conventional recruiting. 

Training that didn’t stop at onboarding 

As the project unfolded, we identified the opportunity to expand the roles. Trainers, many of whom came from OEM or automotive backgrounds, could serve a dual function. They both led the learning program and acted as the primary QA team, reviewing video recordings to perform QA checks and identify gaps in the model, or any safety risks in real time. 

Because the same people responsible for training were also watching performance on the ground, we caught gaps early and updated materials continuously.  

The results: Performance that unlocked the next phase 

The approach delivered measurable outcomes quickly. Within six months of operations, the Japan program achieved 62%+ fleet optimization — surpassing the U.K., U.S. and Germany markets that have been running for eight, five and two years respectively. The program has recorded zero at-fault claims since the start of operations, runs seven days a week and has scaled by more than 300% on the strength of its performance. 

Read our comprehensive case study: Deploying Autonomous Fleets on the World’s Most Complex Roads. 

Dive Brief:

• Stellantis announced a strategic technology partnership with U.K.-based self-driving startup Wayve to integrate the company’s AI Driver into the automaker’s STLA AutoDrive platform, per a May 21 press release.
• The initial integration with Stellantis’ AutoDrive platform will support Level 2+ door-to-door supervised and hands-free driving for both highways and urban environments and is targeted to launch in North America in 2028. Early development on Stellantis vehicles has already started, per the release.
• “This agreement marks an important next step for Wayve and Stellantis in scaling our technology together,” said Alex Kendall, co-founder and CEO of Wayve, in a statement. “By combining Wayve’s AI with Stellantis’ engineering expertise, scale, and global reach, we’re accelerating our mission to bring autonomy to any vehicle, anywhere.”

Dive Insight:

The expanded collaboration will combine Wayve’s end‑to‑end autonomous driving technology with Stellantis’ engineering expertise and manufacturing scale, per the release. 

It marks the next phase of the partnership between the companies following Stellantis’ recent strategic investment in the startup. In February, Wayve closed on a $1.2B Series D funding round with participation from Stellantis, Mercedes-Benz, Nissan Motor Co., Microsoft, Nvidia and Uber Technologies.

Wayve’s AI Driver software stack is designed for universal compatibility, allowing OEMs to more easily add automated driving features to their vehicle portfolios.

According to Stellantis, Wayve’s AI Driver is designed to deliver “human-like driving behavior” that can continuously improve over time by learning from real‑world driving scenarios. The automaker says Wayve’s technology serves as a more natural and more intuitive driving assistant that’s aligned with how its customers actually drive.

“At Stellantis, we focus on technology that fundamentally transforms how our customers interact with their vehicles,” said Ned Curic, Stellantis’ Chief Engineering and Technology Officer, in a statement. “Combining our STLA AutoDrive platform with Wayve’s groundbreaking AI-first approach creates a genuinely intuitive and enjoyable hands-free driving experience.”

Wayve’s AI Driver will also serve as a scalable foundation that can be added to additional Stellantis vehicle platforms and markets over time, supporting higher levels of autonomous driving as the technology evolves using real-world data.

Along with STLA AutoDrive, Stellantis’ core vehicle platform technologies include STLA Brain for onboard computing and STLA SmartCockpit for human-machine interfaces. STLA AutoDrive, the company's first in-house-developed automated driving system, was unveiled in February 2025.

Uber is also working with Wayve to deploy self-driving vehicles on its ride-hailing platform. Uber previously invested in the startup as an extension of a $1.05 billion Series C funding round the company announced in May 2024 that was led by Japan-based SoftBank Group, with participation from Microsoft and Nvidia.

Nissan announced a similar agreement with Wayve in April 2025, tapping the company’s expertise to help build the next generation of its ProPilot driver-assistance system.

Hyundai Motor Group recently shared a behind-the-scenes look around the newly completed Advanced Driving Simulator at its Namyang Research and Development Center in South Korea. The Driving Simulator can reduce months of rigorous vehicle track testing into just hours using a highly accurate computer-simulated environment, according to Hyundai, in a July 13 press release.

Hyundai said it began working on its proprietary virtual validation platform in 2014, but the value of such a platform was not fully realized until the COVID-19 pandemic, when travel restrictions hampered the automaker’s physical vehicle testing both locally and overseas. Hyundai also said it faced challenges as the complex validation process required to test advanced vehicles had outpaced the limits of physical, on-road testing alone.  

Given the widespread adoption of electrification technologies and autonomous driving, plus customer expectations for improved driving dynamics and connectivity, its test engineers faced a paradox. The launch of more advanced vehicle technology requires more physical testing, but the sheer amount of physical prototyping required to conduct these tests can extend vehicle development timelines and budgets.

Hyundai’s solution was to build a Driving Simulator to validate the performance of its vehicles.

The Driving Simulator was developed using the automaker’s own data, software and engineering expertise and took approximately 10 months to complete, from initial construction in April 2025 through to final completion earlier this year, per the release. The simulator also supports co-development through collaborative data-sharing among other global engineering teams for Hyundai Motor and Kia.

Hyundai’s Driving Simulator features a 270-degree curved display and nine 4K projectors. That submerges the engineer in a photorealistic driving environment as if they were behind the wheel of an actual Hyundai or Kia vehicle. It even has a carbon fiber cockpit and production-equivalent steering wheel, seats and pedals for an authentic driving feel, per the release.

The curved displays stream high-resolution road data in real-time without latency to simulate driving on dozens of kilometers of roads.  According to Hyundai, it delivers an accuracy level that test engineers cannot distinguish from actual highway driving.

However, managing the enormous volume of data presented a challenge to Hyundai engineers. Traditional approaches would require loading tens of kilometers of road data into the simulator simultaneously, which would exceed its compute capability, according to Hyundai.

To solve this, Hyundai developed a Terrain Server that streams only the immediate surrounding road data needed for the vehicle’s current position. The proprietary technology reduces latency and delivers ultra-high-resolution rendering by only loading data required for the vehicle’s immediate position. The system can update the terrain information in under a second, enabling seamless operation, according to Hyundai.

“The biggest challenge was building the software environment,” said Pilyoung Jeong, Senior Research Engineer for Hyundai’s Dynamics Functional Concept Development Team, in a statement. “We had to reconfigure the vehicle dynamics analysis tools used at Namyang R&D Center for the simulator environment while ensuring real-time operation of the motion platform without latency. It was a highly complex integration process, but the simulator’s exceptional fidelity in replicating real vehicle behavior made the effort worthwhile.”

The simulator also features linear motion to replicate the driving dynamics of a physical vehicle, including forward, back, left, right, up, down and rotational motion, such as roll, pitch and yaw. Turning the steering wheel causes the entire cockpit to tilt and sway — replicating g-forces, longitudinal acceleration and micro-vibrations that drivers would experience in the real-world. Hyundai engineers also added advanced motion feedback that reproduces micro-vibrations that are imperceptible to humans but are critical for validating vehicle suspension systems and tire performance, per the release.

Hyundai said that recreating this level of sensory realism required a complex mapping effort. The automaker’s engineers precisely scanned its Namyang testing grounds in 1-millimeter increments, capturing every minute road surface detail, including asphalt texture, bumps, slope variations and drainage features. This data was then used to create a highly accurate 1:1 replica map of the physical test track for its Driving Simulator.

What used to take Hyundai engineers days to complete on test tracks can now be validated much faster in simulation, including testing dozens of different vehicle configurations without building a single prototype, per the release. Among the models Hyundai is testing are its N Performance electric vehicles, which include the Ioniq 5 N.

“The most important part is helping with faster decision-making,” said Jeong in a statement. “The simulator lets us verify performance improvements before and after changes easily and quickly, which speeds up decision-making and lets us put more effort into making better cars.”

Dive Brief:

• Lucid is adding hands-free driving functionality — with driver attention still required — on versions of its Gravity electric SUV equipped with its $6,750 DreamDrive 2 Pro package, which includes additional hardware and sensors. The rollout is part of Lucid’s UX 3.6 software update, now available as an over-the-air download. 

• Called Hands-Free Drive Assist, the feature is only available on “compatible highways,” according to a Lucid release. The automaker confirmed to WardsAuto that includes approximately 125,000 miles of roadways across the U.S. and Canada.

• With the update, Lucid is also introducing smarter energy management features, plus better route functionality, with more comprehensive charging-station results and Google Maps Places insights for plotting destinations. The Gravity’s Adaptive Driving Beam headlamps are also now capable of reshaping the high beam around other vehicles automatically.

Dive Insight:

The rollout of Hands-Free Drive Assist on the Gravity starts to realize the potential of the electric SUV’s optional DreamDrive 2 Pro package, which gives the EV a similar set of features rolled out last summer to Lucid Air sedans with DreamDrive Pro.

According to Lucid, the feature will initiate overtaking of slower vehicles, returning to the original lane automatically, or drivers can request a lane change with the turn signal. Lucid says that drivers will be able to choose from Conservative and Assertive driving styles, with the latter offering “a more active approach to overtaking slower vehicles.”

In the Gravity, the automaker says the package includes “future-capable ADAS hardware for driving and parking with one of the most comprehensive sensor suites available,” and it notes plans to roll out an enhanced automated parking feature and green-light alert in future updates to DreamDrive 2 Pro vehicles.

As with the Air with DreamDrive Pro, it adds lidar sensing as part of a 32-sensor array, “to easily navigate traffic and travel with confidence,” according to Lucid. 

The hands-free driving functionality could ease the stress of the commute or the weekend road trip. For comparison to Lucid’s 125,000 miles of U.S.- and Canadian-compatible roads, GM’s Super Cruise now works on over 600,000 miles and Ford’s BlueCruise works on about 130,000 miles.

While that’s an incremental step, it’s an important bridge toward what Lucid sees as a future revenue stream and a key to financial viability. At its Investor Day in March, Lucid said it plans to enable autonomous driving as a monthly subscription in its vehicles starting in 2027. It also previewed a two-seat electric robotaxi concept as a potential advancement of a strategic relationship with Uber Technologies, recently scaled up to include a $500 million investment from the ride-hailing company. The partnership aims for up to 35,000 Gravity SUVs purpose-built for Uber’s ride-hailing network with Nuro’s SAE Level 4 autonomous vehicle tech. 

Lucid, in a February update to investors, had laid out plans to leverage autonomy with SAE Level 3 functionality targeted for 2028 and L4 for 2029, both employing its upcoming midsize platform.

Mercedes-Benz has recalibrated its autonomous driving ambitions, stepping back from an initial plan to offer an updated SAE Level 3 Drive Pilot system in the latest version of its flagship sedan.

For the facelifted 2026 model year S-Class, the German automaker will instead deploy a so-called Level 2++ driver-assistance system in the U.S. market, Mercedes-Benz representatives told WardsAuto over the course of several conversations in the past months. At the same time, it will redirect its engineering resources toward the development of a future Level 3 system capable of operation at speeds up to 81 mph as well as an even more advanced, driverless Level 4 system, which has already begun public road trials in an S-Class-based prototype robotaxi in Abu Dhabi.

The shift comes after a reassessment of both the U.S. regulatory environment and the value proposition for customers, Mercedes-Benz said. Certifying a Level 3 autonomous driving system, where the vehicle assumes legal responsibility for driving under specific conditions, remains a complex, state-by-state challenge in the U.S.

Why Mercedes-Benz is pivoting to its Level 2++ system

Mercedes-Benz’s earlier Level 3 Drive Pilot system, the first of its type to be offered in the U.S., was offered on the S-Class and electric-powered EQS. It was restricted to 40 mph and was geofenced within the states of California and Nevada.

However, Mercedes-Benz has confirmed that its Level 2++ system, known as MB.Drive Assist Pro, will launch across the U.S. later this year in both the facelifted fourth-generation S-Class and the all-new third-generation CLA sedan.

This newer system, which will first be launched in China before being rolled out in the U.S. and, later, Europe, is designed to manage complex urban traffic scenarios, including automated lane changes and navigation in heavy traffic. But it lacks the full fail-operational architecture that defines higher levels of driving autonomy under the SAE’s Level 3 and Level 4 autonomous driving guidelines.

Defending the move, Mercedes-Benz says it will use the new MB.Drive Assist Pro system to advance to Level 3 and Level 4.

“The technical foundation has been laid for the next step,” said Mercedes-Benz spokesman Tobias Mueller.

The sensor suite of Mercedes-Benz’s Level 2++ system, including its lidar sensor and the centralized computing of the latest Mercedes-Benz Operating System (MB.OS) is engineered to be scalable. The transition to a Level 3 system, targeted for inclusion on future Mercedes-Benz models, will primarily require the addition of fully redundant systems for steering, braking and power.

The implications of the decision to roll back on its initial plans for Level 3 autonomous driving extend beyond Mercedes-Benz. It signals a growing caution within the global automotive industry regarding the commercial and regulatory feasibility of incremental, geofenced Level 3 autonomous driving systems in key markets such as North America. The strategy now appears to favor perfecting and widely deploying high-level driver assistance systems while preparing for a more substantial leap to autonomous systems that can achieve highway speeds without human oversight.

In a bid to accelerate development, Mercedes-Benz has entered into technical partnerships in the U.S., including a collaboration with NVIDIA on the core computing architecture and AI software stack necessary for Level 3 and Level 4 autonomous driving. It is also working closely with Uber to explore deployment models for a robotaxi service based on the S-Class.

How the new S-Class compares to previous models

These driver assistance and autonomy plans are grounded in a comprehensive update to the S-Class, some six years after its launch in 2019. The key revisions for the 2026 model are both engineered and stylistic, focusing on drivetrain refinement and chassis intelligence but also subtly updating its looks and cabin.

The most significant powertrain revision centers on the V-8 engine, known internally as the M177, which powers the S 580 4MATIC. The twin-turbocharged 4.0-liter unit has undergone a fundamental change with the adoption of a flat-plane crankshaft and a new firing order. This departure from the traditional cross-plane design is aimed at reducing rotating mass and internal vibrations. Output has risen to 530 hp and 553 lb.-ft. of torque, gains of 96 hp and 37 lb.-ft. over the outgoing model.

The inline-6-cylinder engine, codenamed M256, in the updated S 500 4MATIC has also received attention. Revisions to its electric compressor and camshafts prioritize torque, which now peaks at 443 lb.-ft. While its horsepower rating remains at 443 hp, the focus has been on improving noise, vibration and harshness characteristics. Both the revised, newly engineered V-8 and inline-6 are now paired with a 48V mild-hybrid system featuring a 23-hp integrated starter-generator housed within the 9-speed automatic transmission as standard.

The chassis software has been updated to leverage cloud connectivity. The optional E-Active Body Control suspension can now use anonymized data from other vehicles to preemptively adjust damping for known road surface irregularities.

Exterior design changes are typically restrained, limited to a larger grille and new three-pointed star hood emblem – both with illumination, as well as updated lighting with star-shaped graphics front and rear.

Inside, the primary evolution is digital and centered on Mercedes-Benz’s latest MBUX Superscreen – a new display featuring a standalone 12.3-in. digital instrument display, as well as a 14.4-in. central infotainment touchscreen and a 12.3-in. passenger-side display. The new dashboard-mounted interface is powered by the new MB.OS operating system, which integrates the vehicle's functions in a central computer and introduces a generative AI-powered voice assistant based on large-language models. Other new technology available on the facelifted S-Class includes heated seat belts.

The U.S. remains a critical market for the S-Class, a high-margin sedan flagship in a segment challenged by the dominance of luxury SUVs. Mercedes-Benz’s strategy has been to defend its position not through radical change but through a traditionally evolutionary midlife update, while mapping a more customer-relevant, if delayed, path to more advanced autonomous driving technology, which has proven more complex and taken longer to develop than once thought.

Dive Brief:

• Uber Technologies will invest up to $1.25 billion in electric vehicle company Rivian through 2031 in a deal to add up to 50,000 of the company’s new R2 SUVs to use as robotaxis on the Uber ride-hailing network, Rivian announced in a press release.
• The initial commitment includes the purchase of 10,000 Rivian R2 vehicles by Uber or one of its fleet partners, with the option to negotiate the purchase of 40,000 more in 2030, per the release.
• “We couldn’t be more excited about this partnership with Uber — it will help accelerate our path to level 4 autonomy to create one of the safest and most convenient autonomous platforms in the world, " Rivian founder and CEO RJ Scaringe said in a statement.

Dive Insight:

The fully autonomous fleet of Rivian R2 robotaxis will be available exclusively through the Uber platform. An initial $300 million investment by Uber has already been committed following signing, although the deal is still subject to regulatory approval. 

The full $1.25 billion investment by Uber will be tied to the achievement of certain “autonomous milestones” by specific dates, which were not disclosed. However, if all of the goals are achieved, the companies will deploy tens of thousands of Rivian R2 robotaxis across 25 cities in the U.S., Canada and Europe by the end of 2031. 

Uber’s investment in Rivian follows similar partnerships with Waymo, Nuro and luxury EV maker Lucid Motors that were announced over the past several years.

In September 2024, Uber announced it was expanding its partnership with Waymo to deploy the company’s self-driving vehicles on the ride-hailing platform. The initial U.S. launch of Waymo's electric Jaguar I-Pace SUVs on Uber’s platform was in Austin, Texas, and Atlanta. 

The rollout of the self-driving vehicles on the Uber network was also widely accepted by its customers, according to the company. In May 2025, Uber CEO Dara Khosrowshahi said the average Waymo vehicles in Austin were “busier than 99%” of human-driven Uber vehicles in the city. “Consumers are loving the product,” he said at the time.

In September 2025, Lucid closed on a $300 million investment from Uber as part of a deal to add up to 20,000 of its luxury Gravity SUVs outfitted with autonomous driving technology to Uber’s global ride-hailing network over the next six years. 

In January at CES, Lucid, Uber and autonomous driving technology startup Nuro unveiled a jointly developed “global robotaxi” that was purpose-built for Uber’s ride-hailing network. The EV is equipped with Nuro’s autonomous driving technology and an in-cabin passenger experience designed by Uber.

But Rivian is also developing a proprietary autonomous driving platform and software stack for its vehicles, including the R2. The company in December announced its third-generation autonomy platform, which it claims will be one of the most powerful combinations of sensors and inference compute available in a consumer vehicle in North America. The self-driving technology is expected to be available on Rivian vehicles by later this year. 

Rivian’s third-gen autonomous driving platform includes a sensor suite of 11 cameras, five radars and one lidar unit powered by in-house-developed chips, according to the release.

“We’re big believers in Rivian’s approach—designing the vehicle, compute platform, and software stack together, while maintaining end-to-end control of scaled manufacturing and supply in the U.S.,” Khosrowshahi said in a statement. “That vertical integration, combined with data from their growing consumer vehicle base and experience managing the complexities of commercial fleets, gives us conviction to set these ambitious but achievable targets.”

Rivian recently launched its new R2 SUV, its third EV following the R1T pickup and R1S SUV. In January, The EV maker named Greg Revelle as its chief customer officer to lead its go-to-market strategy for the R2, including sales, marketing and operations. 

The first R2 SUVs are being delivered to wait-list customers this spring. A more affordable version starting at $48,490 is due to arrive in 2027, according to Rivian’s website.

Dive Brief:

• Ford Motor Co. has established a new end-to-end “Product Creation and Industrialization” unit that will work to deliver what it calls one of the most intensive product, software and services rollouts in company history, the automaker announced in a press release.
• By 2030, 90% of Ford’s vehicles by volume will feature updated electrical architectures, in-house-developed hardware, continuous over-the-air software capabilities and improved customer experiences and services, per the release.
• The new organization integrates Ford’s Electric Vehicle, Digital and Design team with its global Industrial System and will be led by COO Kumar Galhotra. “The progress our teams have made in the past few years – from quality and cost to software delivery – has fundamentally reshaped the way we work and positioned Ford for a new era,” said Galhotra in a statement.

Dive Insight:

Ford said its Advanced Electric Vehicle Development team and its “skunkworks” team tasked with designing an affordable, mass-market electric truck has been highly effective, and now the company aims to scale the development model globally and adopt it for future electrified vehicle programs under the new unit.

The Universal Electric Vehicle platform developed by Ford’s skunkworks team features a fully zonal architecture with in-house software controls and advanced driver assist systems that support a wide range of vehicle types and battery chemistries. Ford said the software-based vehicle architecture enables the rapid rollout of the BlueCruise automated driving system and the Ford Digital Experience.

The new Product Creation and Industrialization organization will work to combine these technology breakthroughs with engineering, purchasing and manufacturing. Ford said the unit will be key to achieving an 8% adjusted EBIT margin by 2029.

Alan Clarke will continue to lead Ford’s skunkworks team as Executive Director of Advanced EV Development and was promoted to VP, Advanced Vehicle Development Projects.

“The progress our teams have made in the past few years — from quality and cost to software delivery — has fundamentally reshaped the way we work and positioned Ford for a new era,” Galhotra said in a statement. “By uniting advanced technology with industrial execution, we can make decisions faster, eliminate complexity and deliver great vehicles and digital experiences with the quality and efficiency our customers and shareholders expect.”

Ford also targets a major revamp of its global vehicle portfolio under its Product Creation and Industrialization unit. The automaker aims to refresh 80% of its North American portfolio and 70% of its global portfolio by volume in 2029. New vehicles planned include the first midsize pickup built on its UEV platform, a next-generation F-150 and F-Series Super Duty.

The automaker said its UEV platform development has led to innovations such as more efficient electric motors that will improve the automaker’s future hybrid vehicles. Ford said roughly 90% of its global models will offer electrified powertrains by 2030, including hybrids, extended-range electric vehicles and fully electric vehicles.

“This is the culmination of years of work and progress to create the modern Ford – a talented, unified organization capable of scaling high-quality, software-defined vehicles with a choice of propulsion, distinctive digital experiences and features, and a personalized ownership experience that improves over time,” said Jim Farley, Ford president and CEO, in a statement.

As part of the reorganization, Doug Field, who joined Ford nearly five years ago from Apple and led the development of the Model 3 at Tesla, has elected to leave the company next month, according to Ford. It said in the release that Field helped foster collaboration between the company’s Electric Vehicle, Digital & Design and Industrial System teams that made the integration of the two units possible.

“Doug has been an invaluable partner for me as CEO and helped Ford find its place in this new era of electric propulsion and software-defined vehicles,” Farley said in a statement.

Also departing is Kieran Cahill, Ford VP, manufacturing, Europe and IMG, who is retiring effective May 1 after a 37-year career at the automaker.

Dive Brief:

• At a recent event serving as a first open house for Toyota Motor Corporation’s Woven City mobility test ground, in Japan near Mount Fuji, the automaker revealed its AI Vision Engine, a large-scale AI foundational vision language model that brings together visual, behavioral and environmental data. 
• Toyota claims the AI Vision Engine is capable of identifying patterns, detecting potential risks and enabling coordinated action across connected systems to improve safety. It’s set to be a core building block for the automaker’s next-generation Anzen driver-assist systems and Arene software platform for millions of software-defined vehicles produced in a wide range of configurations.
• Toyota sees commercialization of the AI Vision Engine beyond cars. It employs AI to understand and analyze video data and is being trained from scratch, and Toyota says it might also be applied to retail environments, airports or offices.

Dive Insight:

At the Woven City event, attended by WardsAuto, Toyota further touted the capabilities and versatility of its AI Vision Engine, which was built entirely in-house by Woven by Toyota, the automaker’s software subsidiary. 

The automaker noted in a release that it’s one of the world’s leading VLMs, as ranked by MVBench Leaderboard. And while Toyota’s software unit had cars and mobility in mind when it developed the AI Vision Engine, it sees applications well beyond that.

The AI Vision Engine is capable of understanding the behavior of people, objects and mobility through camera footage, Woven says. But it does not employ facial recognition. What distinguishes it from other VLMs, officials explained at Woven City, is that it can quickly summarize what is happening from camera video, report on what has happened, or help anticipate what will happen. 

“For us, autonomous driving is actually a learning problem,” said Dushyant Wadivkar, Toyota’s global head of AD/ADAS at Woven by Toyota. “It’s less about the algorithm that is specifically programmed for specific scenarios, rather a system that is able to learn and continuously handle the nuanced aspects of driving.”

To solve the learning problem Woven by Toyota is building two core products: a machine learning stack that is designed to learn from drivers, vehicle data and road conditions; as well as an entire ecosystem that powers the stack. Together, they comprise what Wadivkar calls an “active learning loop” that will allow Toyota to extract data from millions of vehicles to train machine learning models, verify and validate data, and deploy it back to customers. 

SDVs that receive regular over-the-air updates define a direct relationship with the end consumer. So with the brand’s image at stake, it’s an ecosystem that Toyota wants to oversee every aspect of, including an execution platform that has to comply with regulatory authorities.

In working toward full SDVs, Toyota plans to set up the platform step-by-step over a period of years. As Woven by Toyota’s CTO John Absmeier explained, the automaker started intentionally with multimedia and ADAS, which he said are the two most complex pieces. 

“But it’s not just about learning for ADAS purposes,” said Absmeier. “We’re talking about the active learning loop for every purpose in the car, so learning about how users use in certain regions, in certain environments, and under certain regulatory and certification differences.”

Of more than 100 million Toyotas on the road around the world, there are dozens of markets for which Toyota distinguishes product lines and model variations differently to deliver vehicles with the right features and pricing. “This creates a tremendous challenge for us to try to provide software to those millions of units that we have to improve over time,” summed Jean-François Campeau, head of Arene and vice president of Woven by Toyota.

To navigate past the build variations and potential incompatibility woes, Arene enters what Absmeier calls a “post-domain era,” with software tools that don’t care about the hardware and can basically test and develop software independently of hardware.

The machine-learning stack is designed to be “generic,” emphasized Wadivkar, who has previously focused on autonomous driving products at Bosch, Cruise and Torc Robotics, so it could be deployed on any car. “We’re not designing a specific application for a specific product, but we’re designing an application system that can work on the multitude of cars that Toyota and Lexus have,” he said.

As part of a vehicle safety system, Wadivkar pointed out that you don’t need a separate application for a five-sensor environment and another for a six-sensor layout, or to redesign the system if using a different lidar supplier.

Absmeier said Toyota plans to deploy Arene globally and across its product lineup “over the coming years,” with the system going into everything from “very, very low-end cars to very, very high-end cars, and in every region of the world.” 

Arene debuted in the 2026 Toyota RAV4 for the vehicle’s infotainment and safety systems. The Lexus ES will arrive later this year with similar deployments of Arene, and Toyota has said its next-generation EVs will be built on it. 

While the active learning loop is going to be fully automated, Toyota at present always has a developer supervising the whole process, according to Wadivkar. Toyota says that although it’s actively testing its Arene-based systems in California, Michigan and Japan, at Woven City it can develop new ideas, and it has willing participants in ”weavers” from the city — people who are there, perhaps with their families, to build things with participating companies. 

Toyota is also tapping its team of professional master drivers normally tasked with product development for help in identifying what “good driving” actually is, which Wadivkar recognizes is challenging and situational. “What is good driving is a very difficult question; but once we get this input from these experienced and highly elite drivers, we are able to curate and use our data very effectively,” he said.

Dive Brief:

• Robotaxi operator Waymo has closed on a $16 billion funding round, bringing its post-money valuation to $126 billion, the company announced in a blog post Monday.
• The funding round was led by Dragoneer Investment Group, DST Global and Sequoia Capital, but also included significant funding from existing investors Andreessen Horowitz, Mubadala Capital, Bessemer Venture Partners, Silver Lake, Tiger Global and T. Rowe Price, per the release.
• The major influx of new capital comes as Waymo continues to scale its commercial autonomous ride-hailing business. The company said it tripled its annual volume to 15 million rides in 2025 and now provides 400,000 paid rides per week in six major U.S. metro areas.

Dive Insight:

Lead investor Dragoneer Investment Group is a San Francisco-based firm that specializes in growth-oriented technology across both public and private markets.

“Dragoneer partners with a small number of companies that can fundamentally transform the world through category-defining technology and products,” said Jared Middleman, partner at Dragoneer, in the release. “Waymo has not only taught a car to drive itself, but to do so meaningfully better than any human or competing system, and we believe that lead will endure.”

Waymo spun out of Google's self-driving car project in December 2016 and is widely considered to be the leading developer of autonomous driving technology with the financial backing of Google (now Alphabet Inc.) over the years. The company also maintains a strong focus on safety. To date, Waymo said its robotaxi fleet has logged over 127 million miles of fully autonomous operation and achieved a 90% reduction in serious injury crashes. 

The company's autonomous driving technology stack installed on its robotaxis is called the “Waymo Driver.” The suite of hardware includes cameras, lidar, radar and an array of external audio receivers to monitor the environment and safely navigate autonomously. The system also uses highly detailed custom maps matched with real-time sensor data to determine a vehicle’s precise location at all times.

With its fresh round of funding, Waymo plans to expand its robotaxi operations in the U.S. and globally, aiming to launch in 20 more cities in 2026, including Tokyo and London, to meet what it calls “exploding global demand for autonomous mobility,” per the release. Waymo’s ride-hailing business is named “Waymo One.” 

“Waymo will boost productivity and accessibility for millions while improving road safety worldwide,” said Saurabh Gupta, co-founder of DST Global, in the release.

In May of last year, Waymo announced a partnership with Tier 1 automotive supplier Magna International to jointly build robotaxis at a new 239,000-square-foot facility in the Metro Phoenix area. The Arizona facility will allow Waymo and Magna to install the company’s Waymo Driver on multiple robotaxi platforms simultaneously and at higher volumes, the company said.

Also last year, Toyota Motor Corp. announced it would collaborate with Waymo on the development of an autonomous vehicle platform, which includes exploring ways to integrate Waymo’s autonomous driving technology into personally owned Toyota vehicles in the future. 

In November, Waymo hired Google executive Steve Fieler as its new CFO, effective Dec. 1, 2025. Fieler previously held the role of business finance officer for Google’s platforms and ecosystems units, and was responsible for products such as Android and the Chrome web browser.

“Steve’s extensive experience will be instrumental in guiding us through this next transformational chapter,” Waymo co-CEO Tekedra Mawakana posted on LinkedIn in November.