The construction-robotics story of the past month has been money for machines. Bedrock raised $600m at a $4.1bn valuation to retrofit excavators, Origin came out of stealth with $20m and a robot that sands walls, Shimizu took a stake in a Sunnyvale humanoid maker, and Korea put state money into a simulator before the autonomous excavator.
Every one of those machines needs something that can stop it. On 22 October, the company that sells that layer raised again.
The Money, Precisely
FORT Robotics, based in Philadelphia, announced approximately $20 million in additional funding led by Tiger Global and Mark Cuban Companies. The headline number is worth taking apart, because it is two things rather than one.
Roughly half is a financing: Tiger Global leading an approximately $10 million round with participation from Mark Cuban Companies, which the release said was expected to close that week. The other half is a further $10 million Tiger Global commitment to the previously announced common-equity PIPE attached to FORT’s pending business combination with Newbury Street II Acquisition Corp — money that arrives with that transaction rather than independently of it. The release puts total committed capital tied to the combination at more than $50 million including earlier PIPE commitments. All of it comes from existing investors; Tiger Global has been on the register since 2022, when it led FORT’s Series B. Prologis Ventures is named among FORT’s backers in the release boilerplate.
The merger itself was announced on 18 August at a pro-forma enterprise value of $556.6 million on a pre-money equity value of $500.0 million, with expected gross proceeds of approximately $201 million assuming no redemptions by Newbury Street II’s public shareholders and roughly $182 million of net cash to the balance sheet after estimated transaction costs. The combined company would be named FORT Robotics Holdings, Inc. and list on Nasdaq as FROB, subject to approvals; Newbury Street II currently trades as NTWO and is led by chief executive Thomas Bushey. Closing was guided to the fourth quarter of 2026. The confidential submission of a draft Form S-4 registration statement to the SEC was announced on 30 September, with the companies saying they intend to file it publicly.
FORT’s subhead frames the listing as a first — “Investors Increase Their Commitments as FORT Plans to Become the First Public Company Dedicated Principally to Physical AI Safety.” That is the company’s characterisation of its own category and is reported here as such.
Griffin Schroeder, a partner at Tiger Global, tied the re-up to the wider wave: “As the opportunity for physical AI continues to expand, we believe FORT plays a critical role in building an independent safety layer that is foundational to its adoption,” he said, describing the firm’s continued investment as reflecting “conviction in FORT’s progress and promise.”
“Automation and autonomy are expanding across nearly every major industry,” said Samuel Reeves, FORT’s founder and chief executive, who framed the increased commitment as giving the company more resources to scale its team and operating capacity.
What FORT Actually Sells
The product is not autonomy. It is the layer that makes somebody else’s autonomy legally and physically survivable near people, and FORT calls it the Trust Layer. The company’s own site breaks it into three parts: human-in-the-loop control, covering both line-of-sight control and offsite remote operation and intervention; Inside-Out Safety, onboard rules logic running over the machine’s own sensors; and Outside-In Safety, rules-based logic drawing on sensors already present in the environment.
Underneath sit conventional-looking hardware and software catalogues — Safe Remote Control Pro, Wireless E-Stop Pro, Nano Safety Controller Pro, Endpoint Controller and Vehicle Safety Controller on one side; FORT Manager, the Rules Engine, Remote Operation and Connectivity Aggregation on the other. That last one was announced on 14 October, bridging cellular, satellite, Wi-Fi and radio links into a single connection, which is a telling thing for a safety vendor to build: a stop command is only as good as the radio carrying it.
On installed scale the release reports more than 19,500 units deployed to over 600 customers, “including Fortune 500 category leaders,” and 25 patents. The six customers it names by name — Agility Robotics, DoorDash, Cobot, Zoox, Textron and Google DeepMind — sit in humanoids, mobile manipulation, delivery, autonomous vehicles, defence and AI research. On 1 October the company announced a strategic partnership with Agility Robotics, extending the two companies’ safety architecture to the Digit 5 humanoid via a new Offboard Safety Bridge.
The Construction Read
Here is the honest measure of relevance, and it is this publication’s judgement rather than the company’s.
Construction is one of nine industries listed on FORT’s industries page, alongside warehousing, defence, manufacturing, lawn care, mining, agriculture, delivery and transportation; the 22 October release’s own boilerplate names a slightly different set, adding energy. The 22 October release names it twice — once among the markets the new capital is meant to serve, together with warehousing and defence, and once among the industries where the solutions are already deployed. FORT’s industries page gives construction a short section headed “Break new ground on safe operation of vehicles and machinery,” pitching safe remote controls with built-in e-stop and drop detection as a way for an operator to run heavy equipment “from a safe vantage point.” No construction customer is named there.
So this is not a construction-AI company in the sense that most of the rounds on this site are. It is a cross-industry safety vendor for which construction is one addressable market among several, and a reader should discount accordingly.
What makes it worth a story anyway is the gap it sits in. Two weeks before this round, NIOSH published a bulletin on safety in construction robotics, covered here on 27 October as the government writing down what makes a construction robot dangerous — a document whose implicit answer is that construction keeps borrowing a safety playbook written for factories, where the machine is caged and the floor does not change. Certified wireless e-stop, onboard collision avoidance that uses the machine’s own sensors, and perimeter logic from site-mounted sensors are, almost line for line, the three things a site needs that a factory cage supplies for free. FORT has been building that for several markets at once. Construction is arriving at the problem with the most money and the least enclosure.
What to Watch
Whether the S-4 is filed publicly, and what it shows by segment. The companies say they intend to file it publicly. If the filing breaks revenue down by industry, that would be the clearest available answer to how much of a safety vendor’s business construction actually is — though the August release’s own disclosures segment by customer size and cohort rather than by sector, so this is a hope rather than a guarantee.
Whether closing holds to the fourth quarter. The guided timing leaves weeks, and $10 million of the money announced on 22 October is contingent on the combination completing. Redemptions by Newbury Street II shareholders would move the $201 million gross-proceeds figure, which the release itself conditions on there being none.
Whether a named construction customer appears in the funding announcements. The 22 October release names six customers, none of them in construction. The longer list in the August merger release reaches further: among the twelve companies it names is Genie, Terex’s aerial-work-platform brand, whose booms, scissor lifts and telehandlers are construction and rental equipment. Whether FORT books that as a construction account is not something either release says.
