Mobility

AI could help build your next iPhone in the Foxconn factory

Machine learning expert Andrew Ng is helping bring image recognition and more to boost quality control.

Building a slide deck, pitch, or presentation? Here are the big takeaways:
  • Former Baidu AI head Andrew Ng is partnering with manufacturer Foxconn to bring AI technologies to the factory floor.
  • The partnership will focus on image recognition for quality control, which could displace human workers.

Machine learning expert Andrew Ng is partnering with Foxconn, the technology manufacturer known for assembling the Apple iPhone, to bring artificial intelligence (AI) to the factory floor, Reuters reported Thursday.

Specifically, Ng's effort will focus on a division of machine learning known as image recognition, which will help with visual inspection and quality control, Reuters reported. The use of machine learning could help improve quality control at the Foxconn factories, but it could also end up replacing some human workers who perform the same tasks—one of the biggest fears surrounding the use of AI in manufacturing.

Being that the machine learning tools are likely cheaper to use than human employees, it could also help with the factory's bottom line and potentially impact the prices of the products that are made there. However, Ng is aware of the potential for worker displacement, Reuters noted, and wants to help train the workers for higher-level jobs that have less of a chance of being automated.

SEE: IT leader's guide to the future of artificial intelligence (Tech Pro Research)

"I would love to help displaced workers gain the skills they need to succeed," Ng told reporters at a press briefing.

The AI technology Ng is offering essentially works as a digital eye to identify any defects in the parts used for assembling products at Foxconn. In one example provided by Ng, a camera with the technology equipped was able to notice a defect in a particular circuit board.

As a form of machine learning, image recognition tools must first be trained on a set of example images before they can begin their own identifications. Typically, these systems can require hundreds or thousands of image samples before they can be used in products, but Ng's system only requires five, Reuters noted.

Being that his image recognition technology requires so few sample images, Ng's services could be quickly adapted to other types of products or even other types of manufacturing.

Foxconn is known as the factory that assembles the iPhone, but it also has a host of other companies and partners for which is manufacturers devices. Ng didn't say which Foxconn customers would be implementing the technology, Reuters reported.

The partnership will come through Ng's new Landing.ai venture, which was also announced Thursday. Ng was the co-founder of Google Brain and worked as the AI lead at tech giant Baidu before resigning in March 2017.

Also see

iphonex.jpg
Image: Sarah Tew/CNET

About Conner Forrest

Conner Forrest is a Senior Editor for TechRepublic. He covers enterprise technology and is interested in the convergence of tech and culture.

Editor's Picks

Free Newsletters, In your Inbox