Honda And Nissan Are Circling A Shared Vehicle OS — And Their Architectures Don’t Match
Neither company has announced anything. There is no joint press release, no filing, no signed agreement on the public record as of this writing. What exists is a stack of official documents from both automakers over the past two years, and if you read them next to each other, the shape of the problem gets clear fast.
Start with what is actually verified. On August 1, 2024, Nissan and Honda announced a joint research agreement covering fundamental technologies for a next-generation software-defined vehicle platform, targeting roughly a year of basic research before deciding whether to take it toward mass production. Four months later they signed a memorandum of understanding on a full business integration under a joint holding company. Eight weeks after that, it was over — the companies confirmed on February 13, 2025 that Honda had proposed restructuring the deal so Honda would be the parent and Nissan the subsidiary through a share exchange, and that the MOU was being terminated. Both sides said the strategic partnership framework would continue.
That framework never died. It’s the container everything since has been poured into.
What A Vehicle OS Actually Is
This is where most coverage of software-defined vehicles goes sideways, so let’s be precise. A vehicle operating system is not the infotainment skin. It’s not the menu structure or the wallpaper. It’s the layer sitting between the silicon and the applications: the hypervisor partitioning one physical processor into isolated virtual machines, a hard real-time OS handling anything safety-critical, a Linux or Android userspace running the cabin experience, and — underneath all of it — AUTOSAR-style middleware talking to the dumb ECUs that still control door locks and window motors.
Related Articles
- Californians Are Turning Against Newsom’s EV Mandate. Someone Should Tell Them It’s Already Dead.
- Rivian Blamed Its Mechanics for a Falling-Apart Rear Suspension. Two New Crashes Say That Wasn’t the Whole Story
Get that layer right and you can ship a new driver-assistance behavior over the air without touching hardware. Get it wrong and every feature request becomes a supplier negotiation.
Honda describes its own OS in exactly those terms: fundamental software controlling the vehicle’s onboard computers and, through them, the entire car.
Two Bets, Two Directions
Honda debuted ASIMO OS at CES 2025 as an in-house vehicle operating system for the 0 Series EVs, alongside a deal with Renesas to co-develop a high-performance system-on-chip for late-decade models. Original, proprietary, Honda’s own.
Then, on May 14, 2026, Honda’s business briefing expanded the mandate considerably. ASIMO OS is no longer an EV-only proposition — it’s going into hybrids too. Honda committed ¥1.0 trillion to software technologies over three years, against ¥4.4 trillion for gasoline and hybrid vehicles and roughly ¥0.8 trillion for EVs. It set a “Triple Half” target of halving development cost, timeframe and workload versus 2025. And critically for the electronics, Honda said it adopted a domain-based E&E architecture.
Nissan went a different way. On December 1, 2025, Nissan announced its Scalable Open Software Platform, a cloud-based development foundation built on AWS, claiming a 75% cut in software test execution time and a unified environment for over 5,000 developers. Then in May 2026 — three days before Honda’s briefing — Nissan and Red Hat confirmed that Red Hat In-Vehicle Operating System would provide the Linux foundation for that platform, with Red Hat engineers embedded in Nissan’s pipeline and the work centered on a next-generation Central Vehicle Computer.
Domain-based versus central compute is not a marketing distinction. It’s a wiring diagram.
Why That Difference Matters Under The Car
A domain architecture groups functions into a handful of controllers — powertrain, body, chassis, ADAS, infotainment — each running its own software with its own supplier relationships, connected over CAN, CAN FD, FlexRay and increasingly automotive Ethernet. It’s incremental. You can evolve it from what you already build. Honda explicitly said it chose this approach for flexibility across country-specific differences and for using outside resources.
A central vehicle computer pushes toward the opposite: a small number of very powerful compute nodes, with zonal controllers positioned physically around the car handling local I/O. The payoff is a shorter, lighter, cheaper wiring harness — one of the heaviest and most labor-intensive assemblies in a modern vehicle — plus a single place to update software.
For anyone who works on cars, the practical consequences diverge sharply. Domain architectures fail in ways technicians already understand: a bad module, a chafed harness on a known circuit, a fault code that maps to a physical box you can price and replace. Centralized architectures concentrate failure. Fewer modules means fewer things to fail, but when the big one goes, you’re not replacing a $340 body control module.
Related Articles
- A Cop Allegedly Ran a Stolen-Car Ring on the Side. Her Badge Wasn’t the Scary Part — Her Database Login Was
- Toyota Takes A Third Of cellcentric: What The Fuel Cell Truck Deal Really Buys
Both approaches also share a headache: replacement modules increasingly arrive blank and require online programming, security credentials and a subscription to reach the OEM’s server. Nissan’s own framing of the Red Hat deal points at the real timeline problem — the platform has to sustain “a mission-critical platform that must remain on the road for decades,” in the words of the company’s SDV software platform manager. Decades of security patches for a car whose original silicon is a museum piece by year eight.
The Corporate Tell
The most revealing sentence in Honda’s May briefing wasn’t about ASIMO OS at all. Under capital allocation, Honda said it will improve investment efficiency by leveraging external resources “without being overly focused on internalization of resources.” The same document lists strategic use of external resources as one of three pillars, calls out industry-standard components, and states Honda will not pursue complete in-house battery sourcing.
That is a company publicly loosening its grip on the not-invented-here reflex, ten weeks before anyone in Tokyo started whispering about a shared operating system. Read Honda’s own words and a joint software foundation stops looking like a surprise and starts looking like a line item.
The math also tracks. Honda’s Triple Half goal — halving development cost, time and workload — is not achievable by working harder. It’s achievable by not writing the same middleware twice.
What A Shared OS Would Not Mean
Here’s where enthusiasts should keep expectations calibrated. A common operating system layer does not produce badge-engineered cars. Steering feel, damper tuning, throttle mapping, ADAS intervention thresholds, HMI design, sound design — those live in application software and calibration files above the OS, and they’re where brands actually differentiate. Two cars can share a kernel and drive nothing alike.
What it would standardize is the boring, expensive plumbing: hypervisor partitioning, safety certification, cybersecurity compliance under UN R155 and R156, OTA delivery infrastructure, and the diagnostic stack. That last one has downstream effects most buyers never think about until they need them.
The Repair And Insurance Angle
Shared software foundations tend to be good news for independent shops and bad news for nobody except tooling vendors. Common diagnostic protocols across two brands means aftermarket scan tool development gets cheaper per vehicle covered, and a wider installed base makes it worth someone’s time to support.
Related Articles
- Toyota And Nissan’s Kyushu Plants Just Got Hit By A Killer Earthquake — Again
- BMW’s New Mexican-Built EVs Arrive In 2027 — Their Batteries Are Still Just Visiting
The insurance side is less rosy regardless of who writes the code. As functions migrate into fewer, more expensive compute modules, the cost to repair a moderate collision climbs, which pushes more vehicles past total-loss thresholds at a given actual cash value. Add mandatory ADAS recalibration after windshield, bumper cover or suspension work — often requiring a static target setup, a level floor and specific lighting — and a fender bender on a software-defined car generates a repair order that would have been unrecognizable in 2015.
If you’re buying into any early SDV, the question worth asking a dealer is not how fast the screen is. It’s how long the manufacturer commits to security updates and whether feature functionality survives if a connectivity service is discontinued.
What To Watch Next
Honda has a habit of filing formal notices with the Tokyo exchange when reports about its plans get ahead of announcements. Nothing has appeared confirming a joint OS. Until something does, the verified position is this: Honda is spending a trillion yen on software while telling shareholders it will lean on outside resources, and Nissan has already outsourced the deepest layer of its stack to an enterprise Linux vendor.
Two companies that couldn’t agree on who would own whom are far more likely to agree on who owns the kernel. Whether Honda’s in-house OS survives that conversation intact is the part nobody has answered on the record.