Insider Brief
- A report from AWS Startups with Strand Partners found that 15% of startups across 20 markets are building physical AI, but customer readiness remains the weakest part of the ecosystem, with just 26% saying their local market is ready to deploy the technology and 70% citing slow adoption as a growth constraint.
- Cloud infrastructure is central to development, with 72% of builders calling it essential, 62% saying they could not have started without it and 62% identifying simulation and synthetic data as critical, while 56% said limited computing power slows progress.
- Commercial deployment is furthest along in manufacturing and automotive at 51%, followed by logistics, warehousing and retail at 47%, while customer demand is driven more often by performance than labor savings overall, though labor shortages lead adoption in markets including Japan, Germany, South Korea and Singapore.
Physical AI startups are moving from prototypes into commercial use, but customer readiness, access to capital and computing resources remain barriers to deployment, according to a new report from AWS Startups with Strand Partners.
The report found that 15% of startups across 20 markets are building physical AI, including robotics, autonomous systems, drones, computer vision, sensors and other technologies that allow machines to operate in the real world.
The findings are based on a survey of 4,420 technology startups, including 663 physical AI builders. Another 16% of startups already use physical AI systems developed by others, 12% are piloting them and 29% are evaluating adoption.
The Cloud as a Key Component
Cloud computing and simulation play a large role in that development. Seventy-two percent of builders said cloud computing is essential to how they build, train or run their systems, while 62% said they could not have started without it. Another 62% called simulation and synthetic data critical, and 56% said limited computing power slows development.
The report found that 60% train models in the cloud, 54% run models on devices for real-time response and 38% use both approaches.
Physical AI Readiness
Labor and skills — 68%. Builders were most confident in their ability to find needed technical and engineering talent.
Supply chains and components — 60%. Most can source needed hardware, although 48% said tariffs or trade barriers restrict access to key components.
Capital and investment — 46%. Fewer than half said they can access the funding needed to build and scale.
Regulation — 40%. Fifty-eight percent said regulatory fragmentation or uncertainty slows development.
Market readiness — 26%. This ranked lowest, while 70% said slow customer adoption is constraining growth.
Deployments
The repoert noted commercial deployment is furthest along in manufacturing and automotive, where 51% of builders said their products are already in use. Logistics, warehousing and retail followed at 47%, agriculture at 40% and healthcare at 37%.
Geographically, the report measured the concentration of physical AI builders rather than commercial deployment rates. Germany led at 24% of startups surveyed, followed by Japan and Israel at 21% and South Korea and the U.S. at 20%. France and Singapore stood at 17% and India at 16%.
Performance Over Labor
Performance, rather than replacing workers, was the most common reason customers adopted physical AI overall, according to the report. Among builders with paying customers, 31% said customers primarily wanted faster or higher-quality work, compared with 22% trying to fill labor shortages and 17% seeking to reduce labor costs.
That balance changes significantly by country. Labor shortages were the leading adoption driver in Japan at 34%, Germany at 30%, and South Korea and Singapore at 29%. Performance led in India at 43%, Vietnam at 40% and Indonesia at 37%, where only 12% to 13% cited worker shortages.
Builders expect adoption to accelerate with 76% saying physical AI will become mainstream in their industry within five years, including 36% who expect it within two years. Three-quarters expect it to reshape their own businesses over the same period.
The report said wider deployment will require greater customer awareness, technical skills, funding, clearer regulation and infrastructure for training, simulation and deployment.
China
China was not included among the 20 markets surveyed, meaning the country comparisons do not show how its physical AI startup base would rank against Germany, Japan, the U.S. and others. The report nevertheless noted that China is investing heavily in physical AI and is already the largest single foreign source of hardware components used by surveyed builders, with 20% naming it as their main supplier. That dependence also creates a supply-chain risk with 48% of builders reporting tariffs or trade barriers restricting access to key components.
