Roborock uses hybrid ToF system by Infineon

The technology allows to reduce size and increase reliability. Compared to the body height of traditional sweeping and mopping robots of around 100mm, the overall height of the Roborock Qrevo Slim is only 82mm, allowing it to pass through lower and narrower spaces while providing high reliability.

“The integration of our REAL3 ToF into Roborock’s cleaning robot is a great example of how Infineon technology is used to accelerate product innovation in the smart home sector,” said Andreas Kopetz, Vice President Ambient Sensing at Infineon. “Our hybrid ToF solution targets a rapidly growing consumer robotic market, enabling customers to develop unique robot designs while reducing system costs and complexity.”

Roborock has teamed up with Infineon, pmdtechnologies, and OMS to develop a next generation robot solution that uses hybrid Time-of-Flight (hToF) to replace the traditional laser distance scanner (LDS) module and obstacle avoidance module. hToF is a combination of Infineon's REAL3 ToF imager, a dual infra-red illumination sources and pmd's processing technology to provide a powerful solution for consumer-grade robots. It supports simultaneous localisation and map building (SLAM), obstacle avoidance, and cliff detection functions in one camera module manufactured by OMS.

“At pmd, we are proud to have played a leading role in the development and optimisation of the hybrid Time-of-Flight (hToF) system that powers Roborock's latest innovation. By contributing our advanced 3D pixels, our expertise in infra-red illumination, and our processing technology into the hToF module, we've been able to create a highly efficient, compact solution that significantly enhances navigation and obstacle avoidance capabilities,” said Jochen Penne, Managing Director at pmdtechnologies ag.

The high-resolution depth and vision data is perfectly suited for robust obstacle avoidance algorithms while the open-source SLAM algorithm based on hToF depth data generates high-precision maps, ensuring accurate and reliable navigation. In addition, hToF is working in darkness and strong sunlight. It is computationally efficient and streamlined, requiring e.g., only one Cortex A55 processor core for depth processing and SLAM computation tasks out of eight available cores.

Roborock has pioneered the mass production of LDS technology since 2016 and is promoting it as the standard configuration for sweeping robots. The company is the first to use the hybrid Time-of-Flight system in their new generation of sweeping robots.