The Capgemini Research Institute has published findings on Physical AI in its report “Physical AI: Taking human-robot collaboration to the next level.” The report describes Physical AI as an evolution in robotics that goes beyond automation, enabling autonomous actions in physical environments and highlighting potential business applications.
The value of Physical AI is recognised across industries, including high tech, warehousing, logistics, and agriculture. The report notes broad agreement among executives in regions such as the US, Europe, and APAC.
The research highlights a period in which technological advances and market conditions are supporting wider deployment of these systems. Developments in foundational models are improving robots’ ability to operate in complex environments. At the same time, progress in simulation technologies is reducing training time by enabling large-scale learning.
The report also describes an AI-robot-data feedback cycle that contributes to ongoing performance improvement and generalisation. Additional enabling factors include edge computing, lower-cost hardware, and business models such as robotics-as-a-service (RaaS), which together support potential expansion of Physical AI.
Most executives surveyed believe Physical AI could enable applications that were previously not feasible, including hazardous tasks and use cases such as dynamic manufacturing assembly and disaster assessment in insurance.
The report links Physical AI adoption to broader industrial trends, including reindustrialisation efforts in Europe and the US. Executives identify reshoring and reindustrialisation as key drivers for adoption, with expectations of increased domestic production supported by autonomous systems.
Workforce dynamics are also highlighted, with labour shortages in sectors such as agriculture, retail, and automotive contributing to interest in Physical AI. The report notes that regions such as Japan are placing particular emphasis on these technologies in their planning.
Beyond addressing labour shortages, Physical AI is associated with operational benefits such as increased flexibility and improved safety, which may support more efficient reconfiguration of production systems.
Despite interest in scaling Physical AI, the report notes that large-scale deployment remains limited. Many organisations report challenges related to readiness. Humanoid robots attract attention but face technical, financial, and training-related barriers, suggesting longer-term development timelines.
The report also notes that societal acceptance is an ongoing consideration, with varying levels of public confidence across regions, which may influence adoption rates.
Overall, Physical AI is presented as a technology with the potential to reshape multiple industries, with its successful deployment dependent on addressing both technical and societal challenges, alongside the development of scalable approaches to human-robot collaboration.
