Tractian raises $130M Series D at $1.4B to scale physical AI for industrial machines

Tractian does not build robots. It builds the nervous system that tells a plant when a pump, chiller or palletiser is about to fail, and investors have just priced that business at $1.4 billion. The Atlanta company said in a PR Newswire release on Oct 22 that it raised a $130 million Series D led by Sapphire Ventures and co-led by M&G's Catalyst strategy. Its MEKAWIRE profile is at Tractian.

The round matters to the robotics market for a simple reason. "Physical AI" money in the US is no longer flowing only to humanoids and robot brains; this month it reached a sensor-and-software company whose product never moves. The question is whether that is a sign of breadth or of a label stretched to fit a fundraising cycle.

The terms, as Tractian stated them

Sapphire Ventures, which also led Tractian's Series C, led again. N47, General Catalyst, Bosch Ventures, 53 Stations, Monashees, Endeavor Catalyst, Norte Ventures and Y Combinator joined as new and existing investors. The company gave the valuation as $1.4 billion and did not say whether that figure is pre- or post-money, nor what the Series C raised. Tractian says it employs more than 700 people, has offices in Monterrey, Sao Paulo and Poland, and that more than 2,000 facilities run its technology. Named customers include Eli Lilly, McKesson, Whirlpool, Coca-Cola bottling operations, Cummins and Embraer, alongside undisclosed data-centre and government operators.

The money, the company said, goes to scaling its multimodal platform and moving from predicting single failures to managing the overall health of every critical machine on a site. That shift is the real product bet.

How the product works

Tractian mounts sensors on rotating equipment and samples vibration, ultrasound and other signals at once. Its models, trained on what the company calls one of the largest industrial failure datasets, turn those readings into a condition picture, a probable root cause and a recommended action. An agentic layer then pushes that recommendation into the maintenance workflow. In practice this is predictive maintenance with a larger dataset and a tighter loop from alert to work order, which is where the company says most value is lost today.

The release leans on two outside numbers: a Siemens estimate that unplanned downtime costs the 500 largest industrial companies about 11% of revenue, roughly $1.4 trillion a year, and a Deloitte projection of more than 2.1 million unfilled US manufacturing jobs by 2030. Both are third-party estimates cited by Tractian, not its own measurements.

Where the round sits in MEKAWIRE's data

Among US physical-AI and robotics rounds of $100 million or more in the last 30 days, Tractian's is the second smallest, yet it is one of only two that came with a stated valuation. The other is SiMa.ai, the edge-AI chip maker, which raised $150 million at $1.45 billion on Sep 28. The two were priced almost identically: the new money is 9.3% of Tractian's stated valuation and 10.3% of SiMa.ai's.