Physical AI is moving from pilot projects to production floors—and Deloitte wants a larger role in that shift.
The consulting giant today announced an expanded collaboration with NVIDIA to build and deploy next-generation physical AI systems, combining high-fidelity digital twins, advanced computer vision, edge computing, and robotics into enterprise-ready solutions. Deloitte is also scaling its global footprint with a new physical AI Center of Excellence (CoE) in Shanghai, aimed at accelerating adoption across manufacturing and industrial sectors.
The message is clear: AI that lives purely in the cloud is no longer enough. Enterprises increasingly want intelligence embedded in machines, factories, vehicles, and physical spaces.
From Digital Insight to Physical Action
“Physical AI” refers to AI systems capable of perceiving, reasoning, and acting in the real world—powering robots, autonomous vehicles, sensor networks, and immersive simulations.
According to Deloitte’s latest State of AI in the Enterprise report, 58% of companies are already using physical AI in some capacity, with adoption expected to reach 80% within two years. That’s a sharp inflection point, particularly in manufacturing, automotive, and life sciences.
Nitin Mittal, Deloitte Global AI leader, framed the shift as operational, not theoretical. Physical AI, he said, is changing how work gets done—on factory floors, in warehouses, and across industrial environments.
By pairing Deloitte’s industry engineering expertise with NVIDIA’s AI stack, the two companies aim to shorten time-to-value, reduce operational risk, and help clients move intelligent systems into full production.
Digital Twins as the Testing Ground
At the heart of the expanded partnership is simulation.
Built on NVIDIA Omniverse libraries, Deloitte engineers are developing immersive digital twins that mirror factories, warehouses, and other complex environments. These virtual replicas allow organizations to simulate workflows, test decisions, and model safety scenarios before deploying changes in the real world.
For example, Deloitte is working with automotive clients to create digital twins of factory and warehouse operations to improve throughput, enhance worker and equipment safety, and cut operational costs.
Simulation-led testing is increasingly seen as a prerequisite for deploying robotics and autonomous systems at scale. It reduces downtime risk and helps validate changes without interrupting live operations—a major cost factor in high-volume production settings.
Edge Robotics and Secure Deployment
Beyond simulation, the collaboration extends into robotics and edge AI.
Deloitte is leveraging NVIDIA Isaac Sim and Cosmos world foundation models to develop embodied AI systems, alongside NVIDIA Jetson Thor hardware to synchronize workloads between edge devices and cloud infrastructure.
In life sciences, Deloitte is supporting the rollout of humanoid systems that integrate simulation, synthetic data generation, teleoperation, and sim-to-real validation. The goal: accelerate safe deployment of intelligent machines in regulated environments without compromising compliance or safety.
Edge computing plays a critical role here. By processing data locally, organizations can reduce latency, enhance security, and maintain operational continuity—even in bandwidth-constrained or sensitive environments.
Computer Vision in Action: A Spain Case Study
Computer vision is another pillar of the initiative.
Using NVIDIA Blueprint for video search and summarization (VSS), Cosmos Reason VLM, and NVIDIA Metropolis, Deloitte deploys video analytics AI agents that combine real and synthetic data to improve inspection accuracy and predictive maintenance.
One recent collaboration took place at an automotive plant in Valladolid, Spain, with Horse Powertrain. As part of an operational efficiency project dubbed kAIros, Deloitte deployed anomaly detection algorithms to predict equipment faults and enhance quality assurance processes.
By integrating on-premises supercomputing with NVIDIA’s AI infrastructure, the plant created a framework capable of rolling out additional use cases across departments—demonstrating how simulation and edge AI can reduce downtime and speed up decision cycles.
The Shanghai Center of Excellence
To support global expansion, Deloitte has opened a new physical AI CoE in Shanghai, adding to a growing international network.
The center focuses on manufacturing applications, including industrial robotics, and aims to help clients navigate security and regulatory considerations as they move from prototype to production.
This geographic expansion is notable. China remains a critical hub for manufacturing innovation and robotics deployment, and establishing a local CoE positions Deloitte to tap into both domestic demand and multinational operations in the region.
Competitive Landscape: Consulting Meets Full-Stack AI
Deloitte’s move reflects a broader trend among global consulting firms racing to embed AI deeper into operational infrastructure—not just IT modernization projects.
What differentiates this push is the full-stack integration. NVIDIA provides the hardware, simulation frameworks, and AI models; Deloitte layers on domain-specific engineering, implementation, and regulatory guidance.
Deepu Talla, NVIDIA’s vice president of Robotics and Edge AI, emphasized that enterprises are moving beyond exploration. The combination of NVIDIA’s physical AI platform and Deloitte’s industry experience, he said, provides a scalable path to production through simulation-first approaches.
Why It Matters
Physical AI represents the next frontier in enterprise transformation. While generative AI reshaped digital workflows, physical AI aims to rewire real-world operations—factories that self-optimize, warehouses that self-correct, and robotics systems that learn before they move.
Early implementations suggest simulation-led testing and secure edge AI can materially reduce downtime and improve operational agility. For industries where minutes of disruption translate into millions of dollars, that’s not incremental—it’s strategic.
The expanded Deloitte-NVIDIA collaboration signals that the infrastructure to industrialize physical AI is rapidly maturing. The remaining question isn’t whether organizations will adopt it—but how quickly they can scale it responsibly.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
