Ultra Closes $62M for Robots-as-a-Service, Taps Physical Intelligence

By Amara Osei

Ultra has raised $62 million to scale its robots-as-a-service business, the company confirmed to Fortune. The funding round, disclosed alongside a new partnership with AI research firm Physical Intelligence, will fund expansion of the company's subscription-based robotics operation.

The headline deal term is concrete: $62 million in, plus an AI research tie-up. Fortune described Ultra as a fast-growing player in the RaaS category. That positioning matters: a fast-growing operator with subscription contracts can outpace investors' baseline growth assumptions and justify software-style multiples before reaching profitability.

Fortune's reporting did not name a lead investor, disclose the post-money valuation, or identify Ultra's chief executive. The publication also left Ultra's customer mix unspecified.

What does the $62 million actually cover?

Robots-as-a-service, shortened to RaaS, restructures how automation reaches buyers. Instead of selling a machine outright, providers sell outcomes: pick rates, inspection cycles, or pallet moves, billed monthly or per task. The operator keeps title to the hardware, absorbs maintenance, and pushes software updates across the asset's working life.

Customers get automation without the balance-sheet hit; operators get a multi-year revenue contract. The economics resemble enterprise software. A robot sold as a service spreads hardware cost across a multi-year contract and lifts lifetime revenue per machine well above the sticker price.

For investors who already pay software-style multiples for predictable recurring revenue, the same shape shows up in a fleet of service-based robotics. That math has powered a string of recent RaaS rounds across warehousing, manufacturing, and inspection. As labor costs rise and sensor prices fall, the payback period for a subscription robot has compressed in some customer categories from years to months.

What the public details do not yet show: the round's lead backer, participating funds, and whether the $62 million came as equity, convertible debt, or a mix. Fortune's coverage also did not break out which verticals Ultra serves today, or whether the partnership includes a commercial component or is research-only. Those gaps will narrow as the company ships its first integrated deployments.

Why does the Physical Intelligence tie-up matter?

Physical Intelligence is the named research partner. The lab sits on the AI side of the robotics stack, where foundation models for perception, manipulation, and control remain an active research front. A deployment partnership gives that research a route into real environments, where lighting, parts variation, and customer-specific workflows expose weaknesses that a controlled lab cannot.

For Ultra, the arrangement compresses the timeline from research paper to production feature. Updated models flow forward to deployed units; deployment data, in turn, flows back to the lab. The bidirectional loop is becoming standard practice across commercial robotics.

The partnership lands during a broader rotation in AI capital. Investor dollars that poured into large language model developers in 2023 and 2024 have begun drifting toward physical AI. Cheaper lidar and depth cameras, maturing vision-language-action models, and tighter labor markets in logistics and manufacturing have created a clearer commercial pull. Robotics operators that combine a working fleet with a research arm are positioned to capture both ends of that cycle.

The near-term product cadence should center on software. Expect closer integration between Ultra's deployed fleet and partner-developed AI, with new manipulation skills and perception upgrades pushed as remote updates rather than hardware swaps. That update channel is the durable advantage a service operator can sell to customers who would otherwise buy a robot and freeze its capabilities on the day of delivery.

The $62 million gives Ultra runway through the next phase of growth. Whether the company converts that capital into a category-defining fleet, or one of several RaaS operators splitting the market, will hinge on execution over the next two to four quarters. The first integrated deployments from the Physical Intelligence partnership — and the unit economics each new customer produces — will be the cleanest signal.