For most of automotive history, the car you drove off the forecourt was the car you kept. Hardware determined what a vehicle could do, and what it could not do was permanent. SDVs are dismantling that assumption. In this post we will look at how software-defined vehicles are changing the commercial model for OEMs, what technologies make the architecture work, and where the regulatory and competitive landscape is heading — so you can assess how ready your organisation is for the shift.
Understanding Software-Defined Vehicles
Definition and Core Components
A software-defined vehicle is one where software, rather than fixed hardware, controls the core functions of the car. Propulsion, safety systems, driver assistance, infotainment, and connectivity are all managed through software layers that can be updated, reconfigured, or replaced after the vehicle leaves the factory. The hardware provides a platform. The software-defined architecture determines what that platform actually does.
The enabling architecture relies on several components working together: a centralised vehicle compute platform, a high-bandwidth internal network connecting ECUs, a secure OTA update channel, and persistent connectivity to cloud and edge infrastructure. Remove any one of these, and the system loses the flexibility that defines it.
SDV Benefits for Automotive OEMs
The commercial case for software-defined vehicles is straightforward. OEMs that previously had a single revenue moment — the point of sale — now have a product relationship that extends across the vehicle’s lifetime. Features can be unlocked post-delivery. Subscriptions can be offered. Performance can be improved in response to real-world feedback rather than waiting for the next hardware generation.
There is an operational benefit too. Recall campaigns that once required physical intervention can, where software is the root cause, be resolved remotely. European OEM competitiveness is under particular pressure here — Asian and American manufacturers have already moved aggressively on this transition, and the gap is not narrowing quickly enough.
Key Technologies Enabling Software-Defined Vehicles
SDV Integration with Cloud Platforms
Cloud integration is not optional in the software-defined vehicle model — it is structural. Vehicle data needs somewhere to go. Model training, fleet analytics, and feature deployment all depend on a reliable, high-bandwidth connection between the vehicle and cloud infrastructure. The vehicle is, in effect, a remote node in a much larger system.
The connectivity layer that makes this work is the eSIM. Unlike a physical SIM card that locks a vehicle to a single network operator, an eSIM allows remote switching between carriers — ensuring the vehicle maintains the best available connection regardless of geography. For automakers building global fleets, eSIM for connected vehicles is what turns a software-defined architecture into something that actually functions across borders.
OTA update infrastructure sits alongside this. The update channel needs to be secure, verifiable, and resilient — a failed or corrupted update at scale is not a minor inconvenience. The systems managing delivery, versioning, and rollback are as important as the updates themselves.
SDV Regulatory Considerations
Regulation has not stood still while the technology has advanced. UNECE WP.29 regulations, now adopted across the EU, Japan, and South Korea, require OEMs to demonstrate cybersecurity management systems and software update management systems as a condition of type approval. Vehicles that cannot meet these requirements cannot be sold in those markets.
This has compressed timelines significantly. OEMs that treated software-defined architecture as a medium-term aspiration have had to accelerate. The regulatory floor has, in effect, made certain capabilities mandatory — which changes the competitive calculation entirely for those still building toward it.
Future Trends in SDV Adoption
The trajectory is toward greater centralisation and greater intelligence. Distributed ECUs — vehicles can have over 100 today — are being consolidated into fewer, more powerful domain controllers. This simplifies the software architecture and reduces the surface area for security vulnerabilities, but it requires significant investment in new hardware platforms and the software stacks that run on them.
AI is accelerating this shift. Driver assistance systems, predictive maintenance, and in-cabin personalisation all depend on models that improve with data — and data only flows reliably when connectivity does. McKinsey’s future mobility research points consistently to software and data as the primary sources of automotive value creation over the next decade, a view now reflected in how the major OEMs are allocating R&D spend.
The longer-term question is not whether software-defined vehicles become the norm — that outcome is already determined — but which OEMs build the connectivity infrastructure to take full advantage. The product is no longer the car at handover. It is the car across its entire operational life, updated, improved, and monetised through a connectivity layer that either works or does not.
Explore how Cubic3 helps OEMs manage connected vehicle fleets.
