InOrbit.AI, robot operations software provider, is leveraging Nvidia’s artificial intelligence (AI) and simulation technologies in its InOrbit Space Intelligence platform.
By integrating NVidia's hardware and software, including the NVidia Blackwell platform and NVidia Omniverse, InOrbit is using physical AI to transform robot fleet management. Applying generative AI, neural networks and reinforcement learning, InOrbit can create self-updating digital twins of any type of robot and the space in which they operate, enabling simulation of business requirements to train AI agents. These agents can then orchestrate any number of robots, pre-playing missions in a simulation and then executing these missions in real life.
Customers like Genentech, a member of the Roche Group, are using multi-vendor robots orchestrated by InOrbit to affect efficiency, accuracy and flexibility.
“InOrbit elevates our robotic deployments from standalone, human-dependent systems to fully embedded automation systems. As we scale our fleet, the role of simulation and optimization becomes even more critical,” shared Matthew Scholten, senior principal robotics engineer at Genentech. "InOrbit provides a complete platform, integrates seamlessly with our existing systems, provides control, performance monitoring and maximizes the value of our robotic deployments."
In partnership with Globant, a digitally native company that helps organizations reinvent themselves and unleash their potential, and their Digital Twin Studio, as well as Dalus, developer of a reliability testing platform for advanced robots, InOrbit is making 3D modeling and simulation accessible to enterprise customers.
"We are thrilled to leverage NVidia's powerful technologies to deliver a new era of efficiency and productivity for industrial customers like Genentech who are integrating smart robots into their workflows," Florian Pestoni, CEO of InOrbit announced. "With InOrbit Space Intelligence, we are using rich data and Physical AI to help companies optimize their robotics investments and launch their facilities into the future."
InOrbit Space Intelligence is designed to accomplish several things:
- InOrbit uses AI to generate a semantic model of a customer’s physical space and to train agents that direct the flow of robot traffic, optimizing workflows and maximizing productivity in their own facility.
- Businesses can have a real-time map of the facility with semantic information. This enables dynamic planning.
- Going from actual robot data in the real environment to agents that drive the robot and using NVidia Omniverse for 3D reconstruction of actual robot missions enables a feedback loop of continuous improvement and higher robot utilization.
"NVidia is providing the building blocks for the next generation of physical AI," emphasized Paul Gutsinger, director of developer relations for NVidia Omniverse. "Integrating NVidia Omniverse and Isaac Sim into InOrbit Space Intelligence showcases the transformative power of AI and simulation, optimizing real-world robot operations and delivering new levels of efficiency and productivity inside facilities."
