Robots steal the show at Big Data Expo

Locally produced robots drew large crowds at the China International Big Data Industry Expo in Guizhou, which ended on Saturday, as the province's growing computing industry is poised to accelerate development in the robotics sector.

"We have achieved panoramic visual perception and can autonomously avoid obstacles in three-dimensional environments thanks to our long-term work in the visual field," said Guo Di, R&D director at VMR Technology Co.

Based in Gui'an New Area — now a national computing powerhouse — VMR markets robots designed for manufacturing and logistics and focuses on machine vision. The company's newly released OmniHead module, paired with 3D sensing, stitches together a panoramic image that gives humanoid robots the information they need for autonomous operation. 

"Many companies sell robot components, such as joints and chassis, but few have tackled the head module. We are already in talks with multiple partners," Guo said.

VMR moved to Gui'an at the end of last year to tap the area's computing power potential. On Aug 18, the firm unveiled OmniHead, a humanoid head module with extensive open interfaces that can fuse visual, voice and other data streams. It can be adapted to a variety of humanoid platforms.

"Globally, the robotics industry is moving toward embodied intelligence. That maturation requires vast computing resources for training, and Gui'an has that capacity," Guo said.

As data and AI industries in Gui'an mature, more high-tech manufacturers are settling there. To date, the zone has attracted 26 data centers. The Gui'an data center cluster now exceeds 90 EFLOPs of computing capacity, with intelligent computing accounting for more than 98 percent of that.

One EFLOP represents a supercomputing performance of 1 quintillion floating-point operations per second.

Projects from tech firms including Huawei, Apple, Tencent and NetEase are proceeding. Officials said they will build a data ecosystem centered on high-quality data sets, develop industry-scale AI models, promote real-world applications and drive growth in advanced manufacturing.