AI in the Physical World: Looking ahead to CES

The annual Consumer Electronics Show (CES) is approaching, taking place January 6th to January 10th in Las Vegas. Breakthroughs in automation and AI have positioned 2026 to be a big year for embodied intelligence, next-generation robotics, and autonomous vehicles. Ahead of CES, Senior Partner Philipp Kampshoff and Partner Ani Kelkar shared perspectives during a McKinsey Live event on advancements in robotics and automation and the potential for these technologies to go further.

Already, AI has allowed robots to be more dynamic and effective. Advancements in software enable robots to learn and adapt to environments in real time, and new hardware can sense and interact with the world more acutely. Robots and other autonomous machines can now learn how to interpret sensory data and adjust movements immediately to respond to many scenarios rather than being confined to narrowly defined, repetitive tasks, which broadens automation opportunities significantly.

Autonomous mobility, including humanoid robots and autonomous vehicles (AVs), is an area with ample potential, thanks to cost improvements, compute infrastructure breakthroughs, and at-scale deployment. To continue to scale autonomous mobility, three factors are vital:

1. Safety. Above all else, safety is imperative to get right. Without physical barriers, advanced sensing, control, and fail-safe systems are integral for humanoids to operate alongside people. Autonomous machines have already shown promise in reducing human error—AVs have reduced car crashes and injuries, for example. Companies can refine these safety measures across all use cases to scale adoption and build trust.
2. Affordability. Current autonomous systems are quite expensive, ranging between $150,000 to $500,000. Reducing costs will become more important in near future, requiring companies to rethink the hardware stack and manufacturing approaches.
3. Accessibility and acceptance. To scale autonomous technologies fully, they must work for everyone. Inclusive designs, business designs that lower the barrier to entry, and integrations with public transit will help expose communities to these technologies and encourage them to try it.

The scope and impetus for automation are increasing as labor gaps persist across industries, spurring investments and picking up momentum. Of course, capturing these opportunities and maximizing the value of these technologies also requires companies to train their workforces thoroughly and structure workflows in a way that safely integrates physical AI into daily operations.

Q&A from the session

1. What actions can cities and governments take to build the “invisible infrastructure” required for large-scale adoption of autonomous vehicles—across zoning, building codes, curb management, regulations, and supporting assets such as parking facilities and depots?

To enable large-scale adoption of autonomous vehicles (AVs), cities and governments will need to build the “invisible infrastructure” that allows AV ecosystems to operate efficiently. McKinsey research shows that policy readiness and urban design are often greater constraints to AV deployment than technology itself. Priority actions include clearly defining and signaling pick-up and drop-off (PUDO) zones in dense areas, modernizing zoning and curb regulations to support dynamic access and reduced parking demand, and ensuring AV operators deploy scalable remote assistance capabilities to handle edge cases and disruptions with minimal network impact.

Equally important is proactive public–private collaboration to align AV supply with urban demand. Cities can work with operators to anticipate peak usage and incentivize shared autonomous services, such as roboshuttles, which McKinsey analysis suggests can reduce congestion. Governments can also facilitate access to strategically located fleet depots and charging hubs—potentially by repurposing underutilized real estate or co-locating with existing public transport infrastructure—and promote equitable access by incentivizing accessible AV fleets for vulnerable populations.

2. Do you expect strategic alliances between AI startups and automotive software players to become critical for accelerating time to market for autonomous and software-defined vehicles (SDVs)?

Strategic partnerships between AI startups and automotive software companies are not only becoming increasingly critical, but they are also already reshaping the pace at which new capabilities reach the market. The pace and complexity of innovation in autonomy, in-vehicle intelligence, and electrification are increasingly exceeding what any single organization can deliver on its own. As a result, OEMs and Tier 1 suppliers are turning to specialized AI partners to access best-in-class capabilities, shorten development cycles, and reduce execution risk.

AI-native companies are enabling advanced voice, conversational, and in-cabin intelligence at a speed that would be difficult to replicate in-house, while platform players provide scalable compute and software stacks spanning infotainment, ADAS, and autonomous driving. McKinsey analysis indicates that such ecosystem-based models—combining OEM scale with startup agility—can significantly reduce time to market, improve software quality, and help automakers focus internal resources on system integration, safety, and differentiation.

***

For more on this topic, see the report Agents, robots, and us: Skill partnerships in the age of AI and the articles “A leap in automation: The new technology behind general-purpose robots,” “Beyond the wheel: Perspectives on autonomous vehicles,” and “Humanoid robots: Crossing the chasm from concept to commercial reality.”