Uber buys Segments.ai to supercharge robotics data

Most people still see Uber as a ride-hailing and food delivery app, but under the hood, it has been running a massive real-world data engine for years.

Now, the tech giant is doubling down on selling that infrastructure to the rest of the tech world. Uber AI Solutions, the company's dedicated AI data division, just acquired Segments.ai to ramp up its lidar and multi-sensor annotation capabilities.

A toolkit built for robots and tractors

Segments.ai was founded in 2020 by Otto Debals and Bert De Brabandere. The founders and their entire team are now moving over to Uber AI Solutions to continue building labeling tools for robotics and autonomous machines.

Their portfolio spans well beyond city street taxis. Over the last few years, Segments.ai built annotation pipelines for self-driving trucks, autonomous delivery drones, and even self-driving farm tractors.

Handling dense 3D lidar point clouds and synchronizing them with camera feeds is notoriously tedious. That niche expertise is exactly what Uber wanted.

Uber noted that it has spent nearly a decade developing internal labeling systems on physical-world data to power safety, autonomy, and AI breakthroughs.

Selling internal tools to 30 countries

Uber AI Solutions essentially packages the proprietary data infrastructure Uber built for its own fleet operations and offers it to third parties.

Back in June, the unit expanded its lineup and opened access to commercial AI labs and enterprises across 30 countries.

The catalog is surprisingly wide. Uber offers access to a global talent network featuring specialists in software engineering, finance, law, and linguistics for complex multimodal evaluation.

It also licenses foundational datasets for generative AI, mapping, and voice systems, alongside task simulation pipelines and testing environments designed specifically to train autonomous AI agents.

As Megha Yethadka, head of Uber AI Solutions, put it earlier this year, the mission is to combine Uber's platform, workforce, and systems to help other teams ship smarter models faster.

Why it matters

We think this is a shrewd play to monetize what used to be a massive cost center.

Building physical-world autonomous systems has burned billions across the tech sector. By productizing its internal pipeline, Uber is turning years of operational grind into high-margin enterprise data services.

For the broader AI ecosystem, the spotlight is aggressively shifting from pure text LLMs to embodied AI and physical robotics. That shift creates a massive bottleneck around multi-sensor and 3D point cloud labeling, putting Uber on a direct collision course with incumbent data vendors like Scale AI.

If you run an autonomous robotics or drone startup, having Uber back Segments.ai could mean more robust enterprise-grade 3D labeling tools out of the box. That said, enterprise acquisitions often risk slowing down small, agile tooling teams, so whether Segments.ai retains its edge inside a corporate giant remains to be seen.