Where Robots Are Made: Japan Retakes the Lead From China
Japan now supplies 36.9% of the 425 robots in our database, edging past China's 32.7% and reversing an earlier China lead. Two countries still cover 69.6%.
Japan now supplies 36.9% of the 425 robots in the Industrial Robotics Hub database, edging past China’s 32.7% for the first time since we started tracking country of origin. A few months ago it was the other way around, China led. The flip is small in percentage-point terms (4.2 points) but it is real, it happened fast, and it traces back to a specific batch of catalog additions rather than anything mysterious. This is not a claim about the real-world robotics market. It is what 425 robots in one database, resolved by brand HQ, actually show.
Which country makes the most robots in our database?
Every one of the 425 robots resolves to a country of origin: 224 (52.7%) carry their own explicit countryOfOrigin field, the other 201 fall back to their brand’s home country, the same resolution logic the site itself runs (a robot’s own field, or its brand’s country, never blank). Seven countries cover the whole catalog, and the top two account for more of it than the other five combined several times over.
| Country | Robots | Share | Brands (robot count) |
|---|---|---|---|
| Japan | 157 | 36.9% | Yaskawa (46), FANUC (38), Kawasaki (24), Epson (21), Omron (17), Mitsubishi (11) |
| China | 139 | 32.7% | Estun (29), Inovance (21), ROKAE (21), Siasun (19), JAKA (18), AUBO (12), Dobot (11), Han’s Robot (8) |
| Switzerland | 62 | 14.6% | ABB (48), Staubli (14) |
| Germany | 29 | 6.8% | KUKA (29) |
| Taiwan | 15 | 3.5% | Techman (15) |
| South Korea | 14 | 3.3% | Doosan (14) |
| Denmark | 9 | 2.1% | Universal Robots (9) |
Country-of-origin mix across all 425 robots in the Industrial Robotics Hub database, resolved from each robot’s own field or its brand’s home country.
Twenty brands, seven countries, and a top two that together account for 69.6% of everything we track. Yaskawa alone (46 robots) is nearly a third of Japan’s total, and FANUC (38) makes up most of the rest. On the China side, no single brand dominates the way Yaskawa does in Japan. Estun is the largest at 29, but the other 110 China-origin robots spread across seven more brands, from cobot specialists like JAKA and AUBO to broader industrial lines like Inovance and ROKAE.
Why did Japan just overtake China?
The mechanism is boring in the best way, which is to say it is not a mystery. irh-catalog-discovery added a batch of Japanese-brand robots to the database on 2026-08-31: FANUC’s new heavyweight, the M-2000iA/2300, and Yaskawa’s entire NEXT series (NEX7, NEX10, NEX20, NEX35), four articulated arms in one family. Every one of those additions is Japan-origin. That single day’s catalog growth is most of what closed the gap.
This is a catalog-composition story, not a real-world market-share claim. We are not saying Japan now builds more robots than China, IFR’s real market data is a different measurement of a different thing entirely, built from actual unit shipments worldwide. What changed here is the mix of the 425 robots we happen to have modeled in our own database, and that mix moved because our own catalog grew unevenly, not because global production shifted.
For reference, back when this column last logged a country snapshot (336 robots, mid-July), China led at 38.1% and Japan trailed at 35.1%, a 3.0-point China lead. Today, at 425 robots, Japan leads by 4.2 points. Both the direction and the scale of the swing come straight from catalog growth, most heavily concentrated in Japanese-brand additions in the back half of August.
Does a robot’s country of origin predict what kind of robot it is?
Here is where the country tags get more useful than a horse race. Look at cobot share by country, the fraction of each country’s robots that are collaborative rather than a fenced industrial arm.
| Country | Cobot share | Cobots / Total |
|---|---|---|
| Denmark | 100.0% | 9 / 9 |
| Taiwan | 100.0% | 15 / 15 |
| South Korea | 92.9% | 13 / 14 |
| China | 42.4% | 59 / 139 |
| Germany | 17.2% | 5 / 29 |
| Switzerland | 14.5% | 9 / 62 |
| Japan | 14.0% | 22 / 157 |
Cobot share of each country’s robot count. Source: Industrial Robotics Hub database, 425 robots, 7 countries.
The three smallest countries by robot count, Denmark, Taiwan, and South Korea, are effectively 100% collaborative. Denmark is entirely Universal Robots, Taiwan is entirely Techman, South Korea is entirely Doosan (plus one palletizer). That is not a national specialization in cobots, it is a single brand wearing a flag. “Country of origin” in those three rows is just brand identity restated.
The two volume leaders tell a different story. Japan and China are both genuinely broad industrial catalogs where cobots are a minority, not the center of gravity. Japan spans articulated arms, SCARA, welding, delta, palletizing, and even all seven of the database’s AMRs (Omron builds every one), and sits at just 14.0% cobot, the lowest of any country we track. China is the more balanced of the two big catalogs at 42.4% cobot, spanning both dedicated cobot brands (JAKA, AUBO, Dobot, Han’s Robot) and broad industrial lines (Estun, Inovance, ROKAE, Siasun) in roughly equal measure. Switzerland (ABB plus Staubli) and Germany (KUKA) both sit heavy-industrial-first at 14-17% cobot, despite ABB having real cobot lines of its own (GoFa, SWIFTI, PoWa).
Put plainly: at low robot counts, country of origin tells you almost nothing except which brand you’re looking at. At high robot counts, it starts correlating with real catalog breadth and product mix, because you’re no longer looking at one brand’s choices, you’re looking at an industrial base.
What does this mean if you’re diversifying suppliers by geography?
Top two countries, Japan and China, cover 69.6% of the whole catalog. Add Switzerland and the top three cover 84.2%. If geographic sourcing risk matters to your buying decision, and this column already covered a real 2026 tariff action that hit both China and Japan-origin robots (Robot Tariffs: China, Japan and the Origin Question), your real option set outside those two countries is thin. Switzerland is the only other country in this catalog with genuine breadth, 62 robots across two brands, both heavy-industrial-first. Everything past that, Germany, Taiwan, South Korea, Denmark, is a single brand wearing a country label. There are exactly three countries in this catalog with more than one brand’s worth of real choice: Japan, China, and Switzerland. Only Switzerland sits outside the two countries this column has already flagged as tariff-exposed.
One honest caveat before you build a sourcing decision on any of this. countryOfOrigin in our database reflects where a brand is headquartered or associated, not a certified customs country-of-origin determination, and it does not track which specific plant a specific unit actually shipped from. The tariffs piece linked above found that KUKA (German HQ), ABB (Swiss HQ), and Kawasaki (Japanese HQ) all run at least some China-based export operations that a real customs declaration would tax at the China rate, not their nameplate country’s rate. These tags are brand-identity geography, genuinely useful for understanding catalog composition and brand strategy, but they are not a substitute for an actual per-unit origin declaration if tariff exposure is the question you’re actually trying to answer. For that nuance, read the tariffs piece rather than treating this post’s country column as a customs form.
The buying-decision reframe is this: a “diversify your supplier geography” strategy sounds like it gives you seven options. In practice it gives you three real ones, and two of those three are the countries most likely to show up on a tariff schedule together. Country of origin, in this database and probably in yours, is less a map of the world and more a map of which twenty brands you’re actually allowed to choose between.
Compare these robots