Uber looks to augment its data labeling expertise with acquisition of Segments.ai

Uber has acquired Belgian data labeling company Segments.ai as a boost to the data labeling division that it launched last November.

The company started the division to help organizations train AI models. It will now be building on Segments.ai’s expertise to refine its own data labeling technology. In particular, it is looking to capitalize on the Belgian company’s strength in LiDAR technology.

In its announcement on LinkedIn, Uber said “ We are excited to welcome Segments.ai to the Uber family. Segments.ai brings solid experience in LiDAR annotation tools, deep expertise in the domain, and an incredible base of clients.”

This isn’t just about labeling, however. LiDAR is an essential component of autonomous vehicle development. It uses laser pulses to determine distances to objects by measuring the time it takes light to return to source and, using this information, creating 3D models of the surrounding environment.

Kathy Lange, research director for IDC’s AI and Automation practice, said that while the acquisition would boost Uber’s strength in the autonomous vehicle market, “it [also] has uses in areas such as weather mapping, various parts of government and in robotics.”

She said that while Uber had only set up the data labeling division relatively recently, it has been using the technology internally for some time. Segments.ai was an attractive acquisition for three reasons: “They had strong technology, they had the talent, and they had the customer base.”

It’s not coincidental that the deal occurs not long after Meta’s acquisition of Scale AI to boost its own data labeling endeavors.  “I see it as a kind of reaction of the Meta and Scale AI deal,” she added. “Taking Scale AI off the table led to a bit of a frenzy in the market.”

The improved modeling from Segments will help their own labeling,” said Tim Law,  research director for AI and Automation at IDC. “They will be looking to integrate the technology to improve safety within their vehicles. It’s all about improving object detection and avoidance.”

As an example, Lange said, the technology will be able to improve vehicles’ performance in the dark. “In those conditions, it can build stronger information on hazards such as other cars or objects in the road.”

But while autonomous vehicles will be the main focus of the acquisition, Uber will also be drawing on the Belgian’s company’s strength in data labeling in general, refining its own offerings and ensuring that it has a more complete technology to offer to its customers. It could mean a new direction for the business, Lange said, allowing them to move into other markets.