Hosting Robot Brains on the Cloud

The anticipated 2020 deployment of 5G networks will effect a tenfold speed increase compared to 4G LTE, creating an environment favourable for the mass integration of various robots into an AI-powered cloud. Beijing-based CloudMinds wants to create such a cloud.

In 2015, Bill Huang left his research position at the China Mobile Research Institute to found CloudMinds. Huang’s business partner Robert Zhang is a telecommunication specialist and the former Head of Service Strategy and Operations for Samsung Telecommunications America. The company’s Chinese name, Data Technology (达闼科技), was inspired by his favorite sentient robot from Star Trek franchise.

Synced spoke with Zhang at last month’s AI World Forum in Toronto, where he participated in the panel discussion The Rise of the Robots. Panelists opined on the future of smart machines in human society, while Zhang advanced his idea of comprehensive cloud-based robot interconnectivity.

Security is the raison d’etre of CloudMinds’ MCS (Mobile-Intranet Cloud Service), an end-to-end system comprising a XaaS cloud, a VBN network based on blockchain technology, and a mobile device for robot control. The network is independent of public networks and therefore immune to cyber attacks.

CloudMinds’ XaaS cloud hosts AI applications such as natural language processing, computer vision and speech recognition that can be accessed by robots. The company also produces its own mobile device for robot control, DATA A1, which can access both the public cloud and CloudMinds’ private cloud; and the wearable META helmet for guiding the visually impaired.

Function pipeline of CloudMind’s Mobile-Intranet Cloud Service

Big ideas need big money, and thus far CloudMinds has raised some US$100 million in funding from SoftBank tycoon Sun Junyi, Foxconn, Keystone Venture, and Walden International. By mid-2016 the company had 180 employees and a budget of US$2 million per month. This may seem extravagant for a startup, but as CEO Bill Huang explains, “a normal startup would not be able to secure the funding needed to enter this market.”

Investors have put both their dollars and their faith into CloudMinds’ high-profile founders and their vision that “by 2025, helpful humanoid robots will be affordable for the average household.”

Zhang says CloudMinds is now searching for suitable markets: “We are looking at certain verticals, such as household vacuuming robots. We plan to start with China and the US, there are lots of business opportunities here, and then move on to Japan and Europe.” The company currently has offices in Silicon Valley, Tokyo, and Beijing, and is looking to establish a base in Canada.

Regarding competition in the IoT cloud market, Zhang says “we fend off competition by doing something none of these companies would be interested in doing. Our system can load Amazon or Google’s public cloud. By maintaining a neutral position we can do benchmarking in search of better technology.”

The rapid growth of cloud computing over the last 20 years has made it possible to integrate mobile computing, wireless networks, cloud computing, and now AI. Zhang says it will take still some time to fully evolve the CloudMind business plan: “The whole robotics market is still changing, but we are pretty clear on the fundamental components of this market and will respond nimbly and flexible.”