Insider Brief
- DataMesh has launched DataMesh Robotics, an embodied AI data product built on its executable industrial digital twin platform to support training, validation, and evaluation of robots in dynamic industrial environments.
- The solution targets a gap between static simulations and real-world operations by enabling industrial processes, events, safety constraints, and business logic to execute and evolve during training, producing task-oriented synthetic data and measurable reward signals.
- DataMesh Robotics is designed to integrate with enterprise robotics stacks, including NVIDIA Isaac Sim and NVIDIA Omniverse, and is currently running pilot programs with industrial partners following prototype validation.
PRESS RELEASE — DataMesh, a leading digital twin and spatial intelligence technology provider, today announced the launch of DataMesh Robotics, an embodied AI data product solution designed for industrial and facilities scenarios. Built on an Executable Industrial Digital Twin, the solution enables robot OEMs and robotics application teams to train, validate, and evaluate embodied AI systems using dynamic industrial environments, industrial-grade synthetic data, and configurable task objectives and reward signals.
As embodied AI moves from research environments into real industrial operations, robotics teams face a growing gap between static simulation worlds and real-world industrial complexity. Industrial tasks unfold over time, follow strict process logic, and are governed by safety constraints, events, and business rules. DataMesh Robotics addresses this gap by enabling training in an industrial world that can run, evolve, and react — not just be visualized.
“At the heart of industrial embodied AI is the need for a training world that changes like the real world,” said Jie Li, CEO of DataMesh. “We go beyond industrial-grade scenes and synthetic data by making the environment executable — so processes evolve, events are triggered, and task objectives become explicit and measurable. DataMesh Robotics aims to become the industrial training environment and data engine for robotics teams.”
From Static Digital Twins to Executable Industrial Environments
Many digital twin solutions today focus on static 3D visualization with real-time data overlays. While effective for monitoring and presentation, such environments are difficult to use for embodied AI training, where robots must operate within changing processes and sequences of constrained actions.
DataMesh’s core capability lies in its Executable Digital Twin, built on the DataMesh FactVerse platform. In this environment:
- Industrial objects can move and interact
- Processes such as manufacturing, inspection, and maintenance can evolve over time
- Events such as alarms, state changes, and task transitions can be triggered
- Business logic and behavior rules can execute during runtime
This dynamic simulation capability enables DataMesh Robotics to generate training data and task feedback that more closely reflects real industrial operating conditions, supporting multi-step tasks with safety constraints and partial observability.
Industrial-Grade Synthetic Data and Task-Oriented Training
DataMesh Robotics provides an end-to-end capability stack covering industrial scene modeling, physical simulation, and scalable synthetic data production. The solution supports multimodal data generation and automated ground-truth labeling for robotics perception, navigation, and manipulation tasks, while also outputting non-visual variables such as process states and operational conditions.
A key focus of DataMesh Robotics is addressing one of the hardest challenges in industrial embodied AI: defining task objectives and reward signals. Industrial tasks often involve strict tolerances, sequential workflows, and safety requirements, making reward design complex and error-prone. DataMesh Robotics offers a configuration-driven approach to defining goals, success conditions, and reward structures, enabling clearer training objectives and more stable learning.
Designed for Integration with Mainstream Robotics Ecosystems
DataMesh Robotics is designed to integrate with modern robotics simulation and training stacks. It supports exporting industrial digital twin assets and data to environments such as NVIDIA Isaac Sim and Omniverse, and fits into enterprise robotics R&D and deployment workflows. The solution supports on-premises, private cloud, and hybrid deployments, with enterprise-grade governance.
DataMesh has been recognized by Gartner® in multiple research reports on Intelligent Simulation, where it was listed as a Tech Innovator and Sample Vendor, reflecting its continued investment in intelligent simulation and spatial digital twin technologies.
Industrial-First Focus and Pilot Programs
DataMesh Robotics primarily serves robot OEMs and robotics application teams working on industrial use cases, including workstation operations, navigation in factories and warehouses, facility inspection and maintenance, and operations in hazardous or restricted environments.
The solution has completed prototype validation and is currently running pilot projects with enterprise partners, including telecom operators and data labeling providers. DataMesh plans to continue expanding its industrial asset library, task templates, and ecosystem integrations.
Image credit: DataMesh
