coScene raises $15M for robot training-data tools
What's the deal? Shanghai-based coSceneDealroom has a profile for this one. Try Dealroom → has raised US$13.8M in a funding round led by Yonggui FundDealroom has a profile for this one. Try Dealroom →, with Tianji, Changzhou Zhirong, and other strategic investors participating. The startup builds data infrastructure for what it calls spatiotemporal intelligence.
What's the endgame? coScene, founded in 2022, makes SaaS tools that collect, monitor, clean, label, and evaluate data for complex real-world applications. Its platform is used in cleaning robots, lawn-mowing robots, low-speed autonomous vehicles, and humanoid robots.
What's it for? The company says the money will support data collection, evaluation environments, and tools for preparing data for robot training.
Why now? China's embodied-AI sector is drawing heavy capital. Reuters reported that companies raised US$12.9B across 322 deals in the first half of 2026 — five times the year-earlier figure.
What could go wrong? coScene's significance will hinge less on the financing headline than on whether it can turn its tools into repeatable customer demand. Reports cite a large overseas data order, but do not establish its economics or delivery at scale.
The signal: Robots need vast amounts of useful real-world interaction data, and that bottleneck is fuelling a "data infrastructure" push across China's robotics market. coScene's round places it among the suppliers betting that the picks-and-shovels layer, not just the robots, will capture value.
Read more: itjuzi.com
Image credit: K3NMA
