Unity Technologies, creator of the real-time 3D development platform, has announced Unity Simulation, a cloud-based simulation product for running multiple instances of a Unity project at scale. Supported by Google Cloud’s infrastructure, Unity Simulation lowers the barrier for anyone to train, test, or validate new products and services in a high-fidelity 3D environment, so creators can bring their ideas to the world faster. Initial use cases for Unity Simulation will focus on computer vision oriented applications for the automotive, gaming, and robotics industries, however, the product can be used for any project, prototype, or concept created in Unity. Unity Simulation is available in closed beta.
“By using some of the latest advancements in Unity, including parts of the Data-Oriented Tech Stack and the High-Definition Render Pipeline, our LGSVL Simulator is able to create digital twins of real-world environments with accurate sensor models to train autonomous vehicles across multiple scenarios.”
Unity is already used for a wide range of simulation tasks, allowing businesses to create virtual environments easily. Unity Simulation leverages the scalability of the cloud to run millions of simulations simultaneously, accelerating time to insight by testing multiple scenarios in parallel. This saves time, manages risk, and can show how products will perform prior to costly public launches.
“In a new era of data-driven artificial intelligence, simulation is the key to unlock the full potential of this game-changing technology. By combining Unity’s real-time 3D development platform with the scale and flexibility of the cloud, Unity Simulation will help businesses accelerate the creation of better, safer, and more reliable products,” said Unity Technologies VP of AI and machine learning Danny Lange.
Train AI in a Safe Virtual Environment
There are some situations where testing and training is mandatory to improve product quality and ensure the system operates safely. Training and validating an autonomous vehicle is one instance where using a virtual environment can dramatically reduce the risk for both the car and those around it, allowing businesses to stress-test systems before they are put on the road.
“By using some of the latest advancements in Unity, including parts of the Data-Oriented Tech Stack and the High-Definition Render Pipeline, our LGSVL Simulator is able to create digital twins of real-world environments with accurate sensor models to train autonomous vehicles across multiple scenarios. By running the LGSVL Simulator on Unity Simulation, autonomous vehicle developers will be able to dramatically accelerate the training of their systems by running multiple scenarios in parallel,” said LG Electronics vice president of engineering, advanced platform lab Seonman Kim.
Test at Massive Scale
Game designers today are inundated with choices about their game, but it can be tedious to assess every option if they can only test one choice at a time. By using Unity Simulation to run cloud-based simulations of their game, developers have a powerful product that can save them time by testing multiple scenarios in parallel.
“By working with Unity to leverage Unity Simulation, our team was able to run different combinations of in-game scenarios to find imbalances early in the development process. We believe Unity Simulation will play a big part in improving player experience in the future as well as help shorten the time it takes to soft launch a new title successfully,” said iLLOGIKA studio director Régis Geoffrion.
Validate Before Going into Production
Unity Simulation allows businesses to validate a new product without a large upfront investment in hardware. In a field such as robotics, businesses can use Unity Simulation to train and validate how a new robot will perform before they build the physical product, allowing them to better plan and design a robot to fit their needs.
Unity Simulation is currently available in closed beta, with new users being approved gradually.