Ropedia, a Singapore-based startup, has landed $22 million in pre-Series A funding to scale data infrastructure built for physical AI, according to DealStreetAsia’s July 24, 2026 report — the latest sign that investors are chasing the picks-and-shovels layer of robotics and embodied-AI training rather than the flashier model layer itself.
Physical AI is the current buzzword for systems that need to perceive and act in the real world — robots, autonomous vehicles, industrial automation — and all of them are bottlenecked by the same thing: high-quality, labeled, real-world data that’s far messier and costlier to collect than text scraped off the web. A $22 million pre-Series A is a meaningful check for a pre-revenue-stage infrastructure play, though DealStreetAsia’s brief item doesn’t name the investors or disclose a valuation, so readers should treat any implied pricing with caution until Ropedia or its backers confirm terms.
Every foundation model needs internet text; every physical AI needs somebody to go out and collect the world.
Singapore has been positioning itself as a regional hub for data-labeling and robotics-adjacent startups, and a raise of this size suggests capital is flowing to that thesis even amid a broader slowdown in traditional data-labeling valuations elsewhere. Watch for whether Ropedia discloses named backers, a valuation, or specific customer verticals — sensor data, simulation, or teleoperation datasets — as it moves toward a Series A.
SG's Ropedia bags $22m pre-Series A to scale data infra for physical AI
