Follow Datanami:
December 14, 2017

AI Heads to Factory Floor

AI pioneer Andrew Ng is launching a startup that would initially focus on revitalizing manufacturing, a strategy that parallels the first enterprise deployments of Internet of Things technologies.

Separately, Ng announced partnership with Chinese electronics manufacturer Foxconn to develop AI technologies and expertise.

Ng, who led artificial intelligence development at Google (NASDAQ: GOOGL) and served as chief scientist at China’s Baidu (NASDAQ: BIDU), said the goal of his startup,, is to extend AI beyond the IT sector and scale the technology in manufacturing settings. Ultimately, he asserted in a blog post, the technology can be used to free users from “repetitive, mental drudgery.”

The focus on manufacturing is seen as a way to scale AI and flesh out amorphous concepts like “digital transformation.” “It is through manufacturing that human creativity goes beyond pixels on a display to become physical objects,” Ng stressed. “By bringing AI to manufacturing, we will deliver a digital transformation to the physical world.”

AI is suited to a variety of manufacturing challenges ranging from capacity management and improving production yields to reducing production costs and other efficiencies, he asserted.

“The strategy of integrating AI — everything from data acquisition, to organizational structure design, to figuring out how to prioritize AI projects — is as complex as the technology, and good AI strategists are even rarer than good AI technologists,” Ng said in announcing the AI startup on Thursday (Dec. 14).

Meanwhile, the partnership with Foxconn (HKG: 2038) gives the AI manufacturing startup “a platform to jointly develop and deploy AI solutions and training globally,” Ng noted.

As automation expands across the factory floor, AI proponents note that manufacturing jobs will require new skills requiring extensive retraining. Ng said his Silicon Valley startup is discussing skills training programs with unnamed partners, including local governments.

The web site notes that it intends to develop technologies and “provide in-house AI solutions” through a variety of manufacturing partnerships. Among the initial offerings are adaptive manufacturing approaches, automated quality control and predictive maintenance. Other AI-based applications include visual inspection systems, calibrating and tuning manufacturing equipment along with “automated issue identification.”

The push to move AI to the factory floor parallels early efforts to forge an industrial Internet of Things in which sensor networks are embedded to streamline manufacturing operations and gauge, for example, when production lines need to be shut down for maintenance.

Ng and others are looking to take factory automation a step further by integrating AI technologies into industrial IoT frameworks.

Among the earliest AI manufacturing applications is visual inspection systems that can outperform human inspectors while operating around the clock. Ng said the startup’s deep learning algorithm was trained on only five images while traditional computer vision systems often must be trained on millions of images.

Besides Foxconn, Ng said his startup is also working with leading global manufacturers to deploy AI systems.

Recent items:

How AI Could Reshape Economies

Exposing AI’s 1% Problem