Crunchbase News published a sector snapshot on 5 November, written by Mary Ann Azevedo, that gives construction robotics a year-to-date number worth sitting with. “Companies spanning construction and robotics have raised about $1.6 billion this year through Oct. 29,” the piece says — “more than 4x the $389 million recorded in all of 2025” and “3x the previous high of $506 million in 2021.” With two months left, the article notes, “funding has already surpassed the previous five years combined.”
That is a real inflection and it deserves the headline it got. But the same article contains the number that complicates it, and it is two paragraphs down: “The sector has recorded 30 rounds this year, compared with 28 in all of 2025.”
Four times the money. Two more deals.
The concentration
Crunchbase is explicit about where the money went. “This year’s surge has so far largely rested on three: Bedrock, Icon and Gravis,” the article says, and together they “accounted for more than $1.3 billion, or about 83%, of the roughly $1.6 billion raised.”
By the article’s own account:
| Company | 2026 rounds | Total |
|---|---|---|
| Bedrock Robotics | $270M Series B (February, led by CapitalG and Valor Atreides AI Fund); $600M Series C (October, co-led by CapitalG and Emergence Capital, at a $4.1bn valuation) | $870M |
| ICON | $65M Series C (March); $200M Series D (October, Baillie Gifford and BlackRock) | $265M |
| Gravis Robotics | $200M Series A (August, led by SoftBank, at a reported $1bn post-money valuation) | $200M |
Five rounds. More than $1.3 billion. Bedrock alone, at $870 million across two 2026 rounds, is more than half the year’s sector total, and more than the $389 million the whole sector raised in 2025 and the $506 million it raised in its previous peak year of 2021 — each taken on its own.
What the other 25 rounds look like
Here the arithmetic is ours, performed on Crunchbase’s published and deliberately rounded figures, so treat it as an order of magnitude rather than a ledger.
If the three concentrators account for five of the thirty rounds and more than $1.3 billion of roughly $1.6 billion, the remaining twenty-five rounds share something on the order of $265 million. That is an average round of roughly $10 million — against a 2025 average, on the same source’s figures, of about $14 million ($389 million across 28 rounds). On these numbers the typical construction-robotics cheque did not get bigger this year. It got slightly smaller, while the headline total quadrupled.
Push one step further. Crunchbase names six other 2026 deals: Reframe Systems’ $40 million in August, led by Energy Impact Partners; Monumental’s $32 million Series B in July, led by Khosla Ventures; August Robotics’ $30 million Series B in May, led by Big Pi Ventures; Gritt’s $26 million Series A in July, led by Obvious Ventures; All3’s $25 million seed in April, led by RTP Global; and Origin’s $20 million in October, which Crunchbase files as a seed co-led by Accel and Zacua Ventures — the company’s own announcement named no round letter. Those six total $173 million.
Which leaves roughly $90 million for the remaining nineteen rounds of the year: an average under $5 million a round, pre-seed and seed scale. Because the $1.6 billion is a rounded figure, that residual is the softest number here — a true total of $1.65 billion would put it nearer $140 million, or about $7 million a round. The conclusion survives either way. Beneath the eleven deals named above, this sector is still funding itself in very small increments.
Analysis: that shape is not a contradiction of the boom; it is what the boom is. Capital has decided which problems it believes are solved enough to industrialise — excavator autonomy via retrofit, and 3D-printed structures — and is funding those at growth scale while the rest of the field stays at the stage where $5 million buys a prototype and a pilot. A sector with a genuinely broad thesis would show a rising deal count. Thirty against twenty-eight is flat.
The retrofit thesis is where the money is
Two of the three concentrators sell the same architectural idea: leave the machine alone and add the intelligence. Bedrock retrofits excavators with a kit that drives them; its own site says “Today, excavators,” and puts “machines of all types” in the future tense. Gravis retrofits existing excavators with autonomy hardware. Between them that is $1.07 billion of the year’s $1.6 billion going into not building a new machine.
We have written about why that keeps winning — Gritt’s pitch is autonomy bolted to a rented skid steer, and Atoms’ acquisition of Pronto was the retrofit thesis being bought rather than funded. The appeal to a contractor is obvious: the fleet is already capitalised, the operators already know the iron, and the purchase decision is a software-ish one rather than a capital-equipment one.
ICON is the exception in the trio, and an instructive one. It is a new-machine company — a printer, not a retrofit — and Crunchbase reports it has raised nearly $800 million since its 2017 founding, has built more than 270 homes and structures across five states and two countries, and expects revenue to grow four times this year. The article quotes an ICON spokesperson on the shift in the question being asked: “Early on, the question was whether a robotic construction system could build a good home,” and now it is “whether a company can deliver at volume, across geographies and project types…” Nine years and the better part of $800 million to get from the first question to the second is a useful calibration for anyone reading a 2026 seed deck.
The exits tell you the category is tradeable
The snapshot also lists the year’s liquidity in “construction robotics and related jobsite automation”: Procore’s roughly $845 million acquisition of DroneDeploy, announced in July and closed in September; Oshkosh’s JLG Industries buying core technology from Canvas in January; HavocAI completing its acquisition of Teleo in March; Atoms completing its acquisition of Pronto in April; and EquipmentShare’s January IPO, which Crunchbase puts at about $747 million gross on 30.5 million shares at $24.50.
Four acquisitions and a listing in ten months is a functioning market for these companies, which matters more to the small end of that funding distribution than the headline total does. A $5 million seed in jobsite autonomy now has visible precedent for where it ends up.
What would change the picture
The demand side is not in question. Crunchbase cites a survey from the AGC and NCCER finding labour shortages the biggest cause of project delays, which is the premise under every deck in the category and has been for years. We have reported the same from the NIOSH safety bulletin and from the deployment gap aired at the Groundworks summit: the constraint is not whether contractors want the labour problem solved.
What would make 2027 a genuinely broad year rather than a concentrated one is the deal count moving. Thirty rounds in a sector this large is still a small number of companies getting funded, and the $1.6 billion headline will be repeated all year without the 30-versus-28 that sits beside it. Both numbers came from the same article on the same day. The second one is the one that tells you how many teams got to keep building.
