Gasgoo Munich- D-Robotics and Bosch Sensortec announced a strategic partnership on July 17. Combining their respective strengths in AI computing platforms and MEMS sensing technology, the two parties aim to build an efficient intelligent ecosystem centered on "integrated sensing and computing" for Physical AI applications.
Specifically, the collaboration leverages DiGu's Xuri S600 flagship computing platform alongside Bosch Sensortec's high-precision MEMS sensors. Together, they plan to deliver reference designs, development kits, and software solutions that address industry-wide pain points like timing drift, temperature drift errors, and multi-domain clock synchronization.
On the hardware front, Bosch Sensortec is providing high-vibration-resistant, low-noise, high-performance inertial measurement units, led by the BMI088. These will be deeply integrated with the Xuri S600's multi-core heterogeneous computing platform. By adopting an innovative "SoC unified master clock" architecture, the system allows the IMU, cameras, motors, and sensors to share a single clock domain—eliminating timing mismatches and long-term drift at the source while significantly boosting motion control and environmental perception accuracy.
On the software side, Bosch Sensortec's underlying sensor drivers, calibration algorithms, and multi-sensor fusion algorithms are being tailored for the Xuri S600 platform. This integration enables a low-latency closed loop across perception, decision-making, and control.
Leveraging D-Robotics's mature AI Native full-stack toolchain, customers can achieve rapid integration, one-click adaptation, and efficient debugging—drastically shortening the cycle from prototype verification to mass production.

Image Credit: D-Robotics
Equipped with a 4-core Nash BPU, the Xuri S600 delivers 560 TOPS of edge-side computing power. Paired with an industrial-grade CPU and real-time MCU architecture, it balances large model inference performance with the real-time control needs of robots, making it suitable for high-load, low-latency scenarios across various embodied devices.
Furthermore, the Xuri S600 has completed on-device optimization for mainstream VLA, VLM, LLM, and voice models. It comes equipped with a proprietary suite of scenario algorithms covering spatial perception, omni-directional positioning, and dynamic motion control. Even with a pure vision approach, it can reconstruct and understand space, navigate complex terrain stably, and adapt to demanding industrial and humanoid operation scenarios.
To date, D-Robotics has partnered with over 20 leading clients—including TARS, Spirit AI, X Square Robot, BeingBeyond, UBTECH, and PaXiniTech. The collaborations span humanoid robots, industrial wheeled dual-arm systems, 3C flexible manufacturing, embodied large models, and multi-modal perception. Multiple co-developed models have already entered phases of real-world testing, iterative adaptation, and mass production verification.
For instance, TARS's A3 robot will integrate the Xuri S600 to achieve large-scale deployment of industrial embodied robots. The Beijing Humanoid Robot Innovation Center's "Tiangong 3.0" has completed computing adaptation based on the Xuri S600, with mass deliveries slated for the second half of 2026. Meanwhile, Humanoid Robot (Shanghai) plans to finish multi-scenario verification for its "Linglong 2.0" by the third quarter of 2026. The Xuri S600 has clearly moved beyond technical verification into a new phase of real-world hardware deployment, scenario validation, and mass production.









