Edge AI Fuels the Rise of Software-Defined Cars

Sonatus unveils an edge AI platform aimed at cutting SDV complexity and unlocking smarter in-vehicle intelligence

11 Feb 2026

At a cavernous hall in Las Vegas, amid screens and concept cars, a quieter shift was under way. At CES 2026 Sonatus, an automotive software firm, unveiled an edge-AI platform meant to sit at the heart of future vehicles. Its ambition is not to add another digital feature, but to reorganise how cars think.

Modern vehicles are already computers on wheels. Yet many still rely on separate hardware modules for each new function, from driver assistance to battery management. This patchwork raises costs and complicates updates. Sonatus proposes a different model: a centralised intelligence layer inside the car, capable of running multiple artificial-intelligence applications at once.

The appeal is straightforward. Fewer components should mean less engineering complexity and lower expense. Software could be updated over the air, allowing new functions to be deployed without swapping hardware. In theory, cars would improve over time, much like smartphones.

On the show floor the firm demonstrated predictive tyre monitoring developed with Michelin and AI-powered diagnostics integrated into a Nissan electric-vehicle platform. The systems are designed to spot irregularities early, giving drivers or fleet operators warning before minor faults become costly repairs. Such use cases are modest, but they hint at broader possibilities.

As vehicles become more connected and software-defined, competitive advantage is shifting from mechanical prowess to digital agility. Carmakers are under pressure to deliver frequent updates and new services while keeping spiralling software complexity in check. A consolidated AI platform could help them respond faster to customer demands and open the door to subscription-based features.

The transition will not be smooth. Automotive software must meet strict safety standards. Greater connectivity enlarges the attack surface for hackers. Regulators are scrutinising how deeply artificial intelligence is embedded in vehicle architecture. A central brain may simplify design, but it also concentrates risk.

Still, momentum is building behind edge AI. What began as pilot projects is edging towards production models. If scalable, update-ready intelligence becomes standard, the car industry’s centre of gravity will tilt further from hardware to code. Those who manage that shift deftly may define the next phase of mobility.