Update: NA Robot Order Value +21% in Q2 — Units Only +4%; Auto OEM −25% While Semi Surges

What changed since the IFR Q3 prelim

Our Q3 industrial robotics update answered the install vintage question: IFR’s April 2026 preliminary results (briefed 24 June) printed a record 621,000 global factory robot installations (+15%), beating the September 2025 575k / +6% forecast, with Asia’s share rising to 79% and the United States flipping from −9% to +11% at 38,000 units. That post sits on an installations ledger. Final World Robotics 2026 still lands 24 September 2026.

This August refresh answers a different delta question with a newer official book: what changed once A3’s Q2 / H1 2026 North American robot order data (released ~11 August 2026) replaces the install-rebound story as the live forward signal? North American companies ordered 8,940 robots worth $622 million in Q2 — +4.3% units but +21.3% order value versus Q2 2025. First-half totals reached 17,995 units and $1.166 billion (+2.0% / +6.6%). Automotive OEM orders fell 25% in H1 while semi & electronics / photonics jumped +38% in Q2. Physical automation is still scaling in North America — but the mix, not the headline unit count, is the news.

The dashboard above is built as an order-book vintage delta: dual-axis units-versus-value path, diverging sector YoY bars, price/mix/bridge delta panels, cobot share fade, ASP×volume scatter, and Q2 customer mix. Use the sector period, industry group, delta lens, and book metric controls to isolate Auto OEM’s pause, the electronics surge, or the ASP step.

The headline table: IFR install rebound vs A3 order book

Read the rows as a lens change, not a restatement. IFR counts robots bolted into factories; A3 counts bookings with member vendors. A strong 2025 install year can coexist with a 2026 order book that grows slowly in units and quickly in dollars. For the 2024 geography levels that still anchor Asia’s dominance, keep the research vintage open; for the flat-year YoY map see the 2026 update.

Value outrunning volume is the structural headline

Toggle the dashboard’s book metric to Value, then ASP. Q2 order value rose 21.3% while units rose 4.3% — roughly a five-to-one growth multiple. Implied average selling price (order value ÷ units) lands near $69,600, versus about $59,800 in Q2 2025 and about $61,200 for full-year 2025. Subtract H1 from Q2 and Q1 2026 looks like ~9,055 units and ~$544M — an ASP near $60,100, essentially flat against the FY 2025 average. The price-mix jump is a Q2 event, not a slow drift.

That matters for operators more than the unit print. Capex cases built on 2025’s ~$61k average — or on older cobot-entry pricing — are already stale for integrator quotes. A machine that costs mid-teens percent more needs either more throughput per cell or a longer payback clock. A3 does not publish payload or sensing breakdowns, so any claim about which features drive the mix is inference. What the data supports is narrower and still material: buyers are committing to materially more expensive equipment per unit.

Detroit paused; the supply base and everyone else did not

The demand-side split is where this release earns a dedicated H2. Automotive OEM orders fell 25% in H1 2026 versus H1 2025 — disclosed as a half-year comparison only; do not quote it as a Q2-only figure. Against that, automotive component suppliers ordered +24% in H1 and +20% in Q2. The assemblers paused; the supply base kept buying. That inverts the 2025 pattern A3 described, when component orders lagged while OEM activity improved.

Interpretation, not A3 text: the timing lines up with a documented EV / program pause wave in Detroit (canceled or delayed body-shop launches remove lumpy welding and material-handling orders without proving a structural exit from automation). The data alone cannot distinguish a capex pause from a secular handoff. What it can say is that North American robotics no longer needs Auto OEM growth to grow revenue in a single quarter — Q2 value rose 21% while that historically largest customer cut H1 orders by a quarter.

Electronics, life sciences, and food absorbed the slack

Sector breadth is the offset. In Q2, semi & electronics / photonics ordered +38% year-over-year; automotive components +20%; food & consumer goods and metals each +18%; life sciences / pharma / biomed +9%. On an H1 lens the same industries look even stronger in places: electronics +35%, life sciences +32%, food +17%, plastics +6%. One honest caveat: life sciences’ H1 +32% decelerated to Q2 +9% — “pharma is absorbing the slack” is an H1 statement, not a Q2 one.

Non-automotive customers accounted for 56% of Q2 units. That is a continuation of A3’s full-year 2025 general-industry majority, not a first crossing of a threshold. The genuinely new development inside automotive is the OEM-down / component-up reversal. Cross-check the global install industry rebound in the Q3 IFR update — electronics led installations at +25% globally in 2025 prelims; the NA order book says that customer class is still accelerating into 2026.

Cobot share is fading while total value climbs

Collaborative robots tell the mix story from the other side. Cobots were 19.6% of FY 2025 units (10.7% of value), 15.4% of H1 2026 units (9.8% of value), and 12.7% of Q2 units (7.1% of value) — 1,137 cobots and $44M in the quarter. Unit share is trending down while total order value climbs. Strip cobots out of Q2 and the remaining ~7,800 units carry about $578M — roughly $74k per machine. Low-cost entry volume is losing mix share even as the two fastest-growing sectors remain cobot-heavy in A3’s commentary (life sciences and electronics), which concentrates competition for that specific supply.

Do not read the fade as “cobots are dead.” Read it as the order book’s center of gravity is migrating toward higher-ASP conventional cells while cobots remain strategically important inside the growth verticals. For robots-per-worker intensity rather than order mix, keep manufacturing robot density beside this update.

Orders are not installations — and the IFR final is still ahead

Two methodology walls separate this post from the Q3 theme print. First, A3 orders ≠ IFR installations. Booking-to-bolt lags typically run quarters; a strong order ASP in Q2 2026 will show up in install tallies later — if at all in the same geography cut. Second, IFR’s 621k / US 38k figures remain preliminary; World Robotics 2026 on 24 September 2026 can revise country and industry ledgers. This update does not replace the Q3 install rebound story. It adds the live North American forward book that capital planners actually quote against while waiting for the final global install print.

Pair the ledgers deliberately: IFR says 2025 was a record install year with a US rebound; A3 says early-2026 NA buyers are still ordering, but Auto OEM is cutting, electronics and components are carrying growth, and dollars are rising faster than arms. Both can be true.

What would rewrite this update

1. A3 Q3 2026 showing ASP reverting toward the ~$60k band would demote Q2’s price-mix jump to a one-quarter spike.
2. Auto OEM orders stabilizing (or components rolling over) would rewrite the pause-versus-handoff debate.
3. IFR World Robotics 2026 revising the 621k prelim materiallyor printing a China 2025 country total far from the ~10× US estimate — would force a global install rewrite on top of this order-book note.
4. A sudden Auto OEM retooling wave (new ICE / EV programs releasing body-shop capex) would refill the historical demand anchor without needing general industry to carry the book.
5. Cobot unit share re-expanding above ~18% while ASP stays elevated would challenge the “higher-value conventional mix” read.

Until those print, the live frame versus the Q3 IFR update is narrow: NA order value is running ahead of units, Auto OEM is in a disclosed H1 drawdown, and electronics / components / food are the growth engines.

Caveats and methodology

1. Orders ≠ installations. A3 reports member-vendor bookings for the United States, Canada, and Mexico. IFR reports installations. Do not splice Q2 order YoY into the 2025 install series without a lens footnote.
2. Implied ASP is derived (value ÷ units). A3 does not publish an official average selling price series; treat ~$69,600 as arithmetic from disclosed totals.
3. Q1 figures used for the ASP path are derived as H1 − Q2 (~9,055 units / ~$544M). They are consistent with the release’s half and quarter but are not a separately published Q1 line in the August note.
4. Automotive OEM −25% is H1-over-H1 only. A3 did not publish a separate Q2 OEM YoY in the cited release.
5. Missing industry cells (metals H1, plastics/other Q2, Auto OEM Q2) are omitted from period toggles rather than invented.
6. Non-auto 56% is a continuation of FY 2025’s general-industry majority, not proof of a brand-new structural regime change on that share alone.
7. Causal stories about EV program cancellations are interpretation layered on timing; A3’s release does not attribute the OEM decline to specific OEMs or platforms.
8. Global IFR prelims (621k, Asia 79%, US 38k) are carried from the prior theme update and remain subject to September revision.

Primary sources: A3 — Robot orders increase in Q2 as automation demand broadens (Aug 2026); IFR US double-digit growth (18 Jun 2026); prior theme baselines in industrial robotics update 2026q3 and industrial robotics update 2026; geography levels in industrial robotics research 2026.

The shareable takeaway

Versus our Q3 IFR prelim, A3’s August order book says North American robotics is still scaling — but Q2 order value rose 21.3% while units rose only 4.3%, Auto OEM H1 orders fell 25%, and semi/electronics jumped 38% in the quarter. Implied ASP stepped up to roughly $70k; cobot unit share faded from 19.6% (FY 2025) to 12.7% (Q2). Physical automation is broadening across industries and getting more expensive per machine, even as Detroit’s assemblers pause. For the install rebound that this order book follows, see the Q3 IFR update; for density intensity see manufacturing robot density.