Robot Safety at Robot Prices

The next time you find yourself on EDGAR, consider using my SECSift reader instead!

It turns SEC filings into clean, readable documents. You can search every word, export anything, merge tables, highlight what matters and gray out the rest, and compare against previous filings.

Not familiar with the reader? Here’s a 3-minute interactive demo (viewable on your computer) and another video I made walking through the entire tool.

🌾 Welcome to StableBread’s Newsletter!

On August 25, China’s Tiangong Ultra humanoid ran 100 meters in 8.86 seconds at the World Humanoid Robot Games in Beijing, 0.72 seconds under Usain Bolt’s world record!

But unlike Bolt, the robot stopped by crashing into a thick mat placed beyond the finish line, before its torso caught fire.

Humanoid robots probably won’t be sprinting through factory floors, but they’ll eventually be working shoulder-to-shoulder with humans.

So how can companies ensure the safety of the humans they employ?

The standard way to make an industrial machine safe is to cut its power the moment something goes wrong. However, IEEE Spectrum points out that doing so “will cause a humanoid robot to fall over,” which is partly why humanoids are still separated from human employees.

Getting humanoids onto the same floor as humans takes a different kind of emergency stop. That’s what FORT Robotics has been solving since 2018.

FORT makes wireless emergency stops, handheld remote controls, and on-machine safety controllers that wire into a robot’s power and motion circuits, so the machine stops even when its own controls fail. Its software manages those devices across a fleet.

The stop doesn’t have to be a power cut. FORT’s controllers can slow a machine instead, or run a delayed stop that lets a humanoid sit down before the power goes (so it doesn’t fall over).

The investor deck’s argument is that every mature machine industry buys safety from specialists, the way carmakers buy brakes from Bosch and Brembo, and that in robotics the only open question is which specialist wins.

If you’ve read my Agility Robotics write-up, you know how early humanoids still are, and the rulebook for working alongside humans in the U.S. is only now being written.

But Europe has already written its rule into law. The EU Machinery Regulation, the safety rulebook every machine sold in Europe has to meet, applies from January 20, 2027. Machines with what the rule calls “self-evolving behavior” (their AI keeps changing how they work) can’t work outside the job and the area they were set up for, have to stay correctable by a human at any time, and have to log every safety decision.

The U.S. has no equivalent rule. Its robot safety standard, ANSI/A3 R15.06-2025, was written for arms bolted to a factory floor, whether behind physical safety barriers or sharing a workspace with a worker, not for humanoids.

The international standard being written for them, ISO 25785-1, is still a committee draft, and its comment period only closed on July 8, 2026.

Aaron Prather, ASTM International’s director of robotics and autonomous systems, said in September 2025 that the finished standards were another 18 to 36 months out, which puts them in 2027 or 2028.

Therefore, a U.S. site that wants humanoids next to its workers (none are allowed to run that way today) selects its own e-stops and safety controllers, since no U.S. standard sets the bar yet. And any U.S. builder shipping into Europe has to meet the EU rule in 2027 regardless.

So just how big is the humanoid robotics market?

FORT’s deck leans on three market-size estimates, all made by other firms for robots and physical AI broadly:

Slide 28 puts them side-by-side:

For scale, FORT’s own revenue was $11.6M in 2025, up 61.6% y/y at a 66.2% gross margin, and it lost $4.6M. Bookings, the orders customers commit to, more than doubled y/y in Q1 2026, from $1.72M to $3.50M.

FORT is going public via a SPAC led by Newbury Street II Acquisition Corp (Nasdaq: NTWO) at a $556.6M pro forma enterprise value, and the combined company will trade as FROB once the merger closes, which both sides expect in Q4 2026.

At $10.93/share on the deck’s 74.2M pro forma shares, the combined company is worth $811.0M ($10.93 × 74.2M), or a $625.2M EV after the $185.8M of cash it expects at close with no redemptions.

The trust held $10.73/share at June 30 and accrues ~$0.09/quarter, so it’ll hover ~$10.90 by a December vote. At $10.93/share you’re paying a 1.2% premium to today’s ~$10.80, and redeeming at a December vote would cost you only 0.3%.

In this 7.5k-word deep dive, I discuss what FORT sells and who it competes with, the nuance behind the 600-customer base headline, where humanoids and Europe’s new rules fit, the SPAC terms, and the sponsor’s record. Then I conclude with what FROB is worth after close and whether I’m buying NTWO before the vote.