Physical AI: the new frontier of human-robot collaboration
A · Strong evidence A Capgemini Research Institute survey of 1,678 director-level executives in companies with over $1 billion revenue across 16 countries, with a published method, though outcomes are expectations rather than measured deployments.
Summary based on the full text·Original: about 90 minutes to read
In short
A 140-page report from the Capgemini Research Institute (April 2026) on "physical AI": robots driven by multimodal foundation models that can adapt to unfamiliar tasks and unstructured environments. It combines a survey of 1,678 executives with expert interviews. The main finding is that adoption is underway and driven by labour shortages, with 27% of organizations deploying or scaling, while humanoid robots remain a longer-term bet held back by reliability, dexterity, cost and public acceptance.
Key numbers
- 79% / 27%organizations engaging with physical AI / deploying or scaling it
- 74%executives citing labour shortages as a driver of adoption
- 7 yearsaverage expected timeline to deploy humanoid robots at scale
- $40.7 billionrecord VC investment in robotics companies in 2025, 9% of all VC funding
As reported by the source.
English·PDF·opens capgemini.com
What it reports
- 67% of executives expect physical AI to transform their industry (US 73%, Europe 64%, APAC 67%); 64% expect it to become a critical driver of competitiveness.
- 79% of organizations are engaging with physical AI, 27% are deploying or scaling, and 65% expect to reach scale within five years.
- Drivers: labour shortages 74%, rising labour costs 69%.
- Expected benefits: productivity 76%, cost efficiency 70%, product quality and precision 65%, operational resilience 57%; 60% expect robotics in areas previously impossible or impractical.
- Humanoids: 67% believe they will transform their industry, 53% invest or plan to, 78% expect eventual scaled deployment, but the average timeline is seven years and 30% see viable general-purpose humanoid workers within three to five years. 62% cite public acceptance as a critical hurdle.
- Humanoid production costs fell roughly 30-fold over a decade, from about $3 million to around $100,000; actuators account for about 50% of production cost.
- Robotics VC reached a record $40.7 billion in 2025; investment in world models rose from $1.4 billion (2024) to $6.9 billion (2025).
How they got it
Online survey in January and February 2026 of 1,678 director-level and above executives in organizations with annual revenue above $1 billion ($500 million for aerospace/defence and public services), across 16 countries and 15 industries, from automotive to agriculture. Survey findings are combined with interviews with experts in robotics and physical AI and with secondary market data.
Limits
- Most figures measure executive expectations and intentions, not installed robots or measured productivity.
- No baseline separates foundation-model robots from conventional automation in the adoption numbers.
- Capgemini sells engineering and integration services for robotics and physical AI; its own innovation chief notes robotics' history of overpromising.
Why it matters for companies
For firms with plants, warehouses or field operations, the report frames robotics as a response to labour scarcity rather than a cost play alone. Its practical advice is conservative: start with dull, dirty or dangerous tasks, redesign handovers and supervision, enforce safety through mechanisms independent of the AI layer, and avoid defaulting to humanoids. The seven-year humanoid timeline is a useful counterweight to vendor demos. Workforce planning should expect mixed human-robot shifts in specific tasks well before general-purpose robots arrive.
Summary based on the full text. Read the original for the full argument and data.Open PDF on capgemini.com
- Tags
- Operations