Building a Connected Metaverse

By extending the power of Physics based Simulation and 3D visualization with Network Simulation at scale, organizations could accelerate their digital journey.

As the Metaverse becomes a reality, organizations are scrambling to figure out their digital strategy around the metaverse. Many companies have started building digital replicas of their physical products. The benefits of a metaverse in the B2C market are already clear. From a brand awareness and marketing perspective, having a presence in the metaverse is beneficial to capture the target audience and increase engagement. However, how could B2B businesses leverage the power of Metaverse? From an industrial IoT perspective, building a product in Metaverse could help develop new features faster with a faster feedback cycle. E.g. BMW used metaverse (see video below) to fine-tune their factory automation floor for maximum productivity and worker comfort. Similarly, Industrial customers could provide feedback on the new version of the product even before the first prototypes are ready, by interacting with the simulated product in the metaverse.?

One of the leading examples of bringing Metaverse into industrial IoT is NVIDIA Omniverse. Together with industrial partners such as Siemens, and BMW, NVIDIA has built the digital equivalent of Train stations, Warehouses, and much more such use cases in development.

Here is an example of how NVIDIA collaborated with Deutsche Bahn to build a virtual model of their trains and stations.?

Another fascinating example of collaboration between BMW and NVIDIA is to build a full-scale digital factory prototype in Omniverse.?

The metaverse could be a great enabler for building your digital twin application, where the physical products do not exist yet. With a realistic simulation of the product under development, the Metaverse could simulate a virtual device and its environment, to accelerate the development of cloud applications. However, in order to truly reflect reality, the data coming from Metaverse needs to be packaged into IoT protocols such as MQTT/Modbus and then fed to the IoT cloud platform. In other words, the Metaverse version of the product could behave exactly like a physical product, and your digital twin application will be built around streaming data from this alternate reality.

At IOTIFY, our core expertise has been creating stateful IoT simulations at scale, enabling Millions of connected virtual IoT endpoints to the cloud platform. Our scalable SaaS platform allows our customers to execute performance test / functional tests on their IoT solutions. However, one of the missing pieces of the equation has been to bring physics into the equation i.e., integrating a 3-D physics engine and object representation into the network simulation environment. E.g., How does the acceleration change when a simulated vehicle collides with another simulated truck? What happens when the load of one simulated object is transferred to another simulated object? Would that automatically affect the simulated state of the objects? The answer could be a scalable physical engine that could mimic not only the network state of the objects but also the simulated physical state of the objects. However, such integration has some challenges.

The cost of?scaling

One of the critical restrictions of building a simulated metaverse is the cost of simulating at a large scale. E.g. A minimum of 8 GB GPU is recommended for running the Issac Robotics simulation. Imagine you would like to simulate a million such connected robots in the Metaverse. It would be prohibitively expensive to run such a large-scale simulation.

Limits of physics?engines

The 3D engines tend to become slower as the complexity of simulation exceeds. Other overheads, such as Memory and disk space requirements, come into the picture when the scale of the simulation becomes excessive. As more and more items get added to the simulation, the timestep becomes larger and simulations become slower, to the point of being completely unusable.

IoTIFY’s vision for Connected Metaverse

At IoTIFY, we have developed a unique approach to the scalability problem in the connected Metaverse. By slightly altering the scope of object interaction from a global to a hyper-local space, we could drastically cut down the CPU and memory requirement of running this simulation at a large scale. Finally, a large-scale hybrid simulation could be achieved, drastically reducing the cost of end-to-end connected metaverse simulation.

An example of how we developed a virtual office space for one of our PoC pilots where we simulated fixed objects as well moving persons in the space.?

The purpose of this simulation was to simulate and test the cloud applications which could manage a smart office environment. Since the applications also rely on data captured by Lidar, we also generated synthetic lidar data from the same 3D space.?

Together with live video streaming, the lidar point cloud data was also streamed to the cloud application to simulate a realistic office space environment. Furthermore, the simulation was replicated at scale to simulate a large-scale system.?

Conclusion: It is clear that connected metaverse could help organizations build their products virtually and iterate faster before they start developing the actual products in the lab. By combining advanced network simulation and firmware emulation with the Metaverse, the product development cycle could even further be sped up.?

In order to learn more about how could IoTIFY help you in your connected metaverse journey, please feel free to reach out to [email protected]