GM maps 2028 autonomous highway rollout

General Motors (GM) plans to introduce highway autonomous driving on the Cadillac Escalade IQ in 2028, establishing the large electric SUV as the launch platform for its next wave of driver-assistance technology. The company intends to follow this debut by extending the capability across major internal combustion full-size SUVs and trucks, three-row crossovers and fourth-generation electric vehicles, including high-volume gas-powered lineups such as full-size SUVs and light-duty pickups.

Pre-production vehicles equipped with the system are already being tested on closed courses. GM describes the initiative as one of its largest multi-segment introductions of software-defined vehicles, with the same technology base designed to operate in both internal combustion and electric models.

The highway autonomous driving feature relies on a new electrical and computing architecture that separates hardware from software and standardizes a computing, networking, software and interface framework across vehicle types. At the core is a liquid-cooled Central Compute Unit designed in-house, which delivers roughly 40 times the computing performance of current vehicles while consolidating software for autonomous driving, active safety, chassis and propulsion into a single high-speed system.

This centralized architecture is designed to support in-vehicle artificial intelligence features and the autonomous driving stack running simultaneously while preserving computing headroom for future functions. Zonal concentrators connected by high-speed Ethernet link sensors and actuators over a high-throughput data network, enabling more data processing than existing systems.

The new setup also reduces wiring length and weight by about 30% and increases cellular connection speeds by up to 50%. These changes are intended to enhance in-vehicle communications and connected services and streamline the rollout of software-defined features across multiple segments.

GM expects the centralized computing platform to accelerate development of software-defined vehicles. Once improvements are validated on one model, they can be propagated across the lineup through over-the-air software updates without re-engineering each individual variant. After the initial deployment on the Cadillac Escalade IQ in 2028, GM plans to extend the architecture to gas-powered full-size SUVs and trucks, three-row crossovers and its fourth-generation EVs.

The architecture is propulsion-agnostic, allowing the same software foundation to operate in a gas pickup or an electric SUV. GM characterizes this as a foundation for energy management, in-cabin features, and autonomous driving capabilities, all orchestrated by the central computing unit.

Development and validation of the autonomous highway system draw on GM’s accumulated experience with Super Cruise and extensive AI training. Super Cruise customers have logged more than 1.2 billion miles of hands-free driving, while GM’s driving stack generates more than three billion virtual miles per day, a volume estimated to match what about 2.5 million vehicles would cover running continuously on real roads.

GM states that it will use this learning process, along with real-world road validation, to verify the safety and responsiveness of its new autonomous capabilities. The company’s broader technology plan, presented at its GM Forward 2026 event in New York, integrates AI-accelerated vehicle development, a fourth-generation EV platform and an expanded OnStar intelligence system with the centralized compute architecture to create vehicles whose capabilities can grow over time via software.