Ubtech Unveils Thinker-Sim Embodied AI Simulation Platform for Robot Capability Evaluation

IT Home news on November 2, UBTECH Robotics today released the embodied intelligence simulation platform Thinker-Sim, constructing a one-stop technical system covering real-scene digital and large-scale simulation data synthesis and robot capability evaluation. Let the real world quickly enter simulation, Verify robot capabilities.

According to the introduction, the platform adopts UBTECH's self-developed Thinker-GS high-fidelity scene reconstruction algorithm and full-process SimReady asset automated generation technology, Achieve from real-scene collection、efficient conversion from 3D reconstruction to simulation asset generation.Based on real factory production line environments to perform 3D reconstruction, it can restore the factory area spatial layout, environmental structure, and visual details, and flexibly load foreground assets such as robots, production equipment, and personnel according to different task requirements such as robot training, data collection, and algorithm evaluation.

IT Home learned from the UBTECH official account that the platform can also, for application scenarios such as commercial services, quickly generate digital twin environments with realistic visual performance, supporting robots to carry out training and deployment verification in different scenarios. It supports technologies such as image and text-driven batch generation of 3D assets, low-poly topology optimization, visual material generation, and automatic collision body configuration, balancing high-quality visual performance, physical simulation adaptation, and large-scale production efficiency.

In addition, The platform has built-in rich standardized simulation task templates, covering common robot training, data synthesis, and evaluation scenarios. Users do not need to build simulation environments from scratch; by selecting templates and configuring parameters, they can quickly start simulation tasks. It supports real-time screen streaming of the simulation process, task status monitoring, and execution result feedback, allowing users to grasp the simulation execution status in real time.