Company history

FUJI began as a refusal. At Suga Iron Works, where he worked before going independent, Sakagami Mamoru had designed a tannōki — a single-purpose machine tool, built to perform one operation rather than many — and urged his employer to put it into real production. He was turned down, and in April 1959 he set up Fuji Machine Mfg. in Nakagawa-ward, Nagoya, with capital of $833 (¥300,000). The choice of the single-purpose machine over the general-purpose lathe fixed the company’s method for the next sixty years: accuracy is not sold off the shelf, it is built into a customer’s own production line.

The first FS-type machine went out in May 1959, and word of its productivity brought buyers to the plant who ordered on the spot. Fuji moved to Chiryu, east of Nagoya, in 1961, and sold automatic lathes to the region’s manufacturers of bearings, electrical parts and motorcycles — and, above all, to the car industry: Toyota, Honda, Mitsubishi and Isuzu. At Toyota, Fuji machines came to occupy more than 70% of the lathe lines. In May 1964, five years after founding, the company listed on the second section of the Nagoya Stock Exchange, and that October raised capital to ¥200 million. Financing the head-office plant had been the hard part — the company was too young to carry much credit — and a young bank officer named Otsuya Teruo, who would later become its second president, helped find the money.

The year after listing, the 1965 recession hit machine-tool demand and Fuji cut about sixty employees. It is the only workforce reduction in the company’s history, before or since; every later crisis was met without one. The episode ended management by daily improvisation: basic policies, profit plans, capital-investment plans and budget control all date from this moment. Business with Toyota was won in the same year, a transfer line for volume machining followed in 1967, and the product line kept widening — an automatic assembly machine and an NC automatic lathe in 1971 — until the 1973 oil crisis pushed the domestic machine-tool market into low growth for good. What the assembly machines had taught, though, was how to position a part precisely and move it quickly. That was the technology the next business would be built on.

In October 1978 Fuji completed an automatic electronic-component inserter — a machine that pushed leaded components into the holes of a printed circuit board, work that televisions, radios and calculators were still having done by hand. The development was ordered by Otsuya Teruo, then running the administrative side, and the notable thing he did was not the order but the waiting: for two years the machine could barely hold 200-micron accuracy and inserted only about nine parts in ten, and the fix came only after the control microcomputer was changed and the mechanism reworked. Some 850 of them were eventually sold — enough to establish the business.

The successor mattered more. In July 1981 came the automatic placement machine, which puts surface-mount components onto the face of a board rather than through it — exactly the direction board design was taking. Its launch was rough: the first CP-series machines generated more trouble at customer sites than the inserters had, and salesmen spent their time on complaints. Inside a company that still regarded machine tools as the real business, the division was a sideline whose economics were openly doubted, and it heard the question of how much longer it intended to go on making assembly machines. It survived on what it had inherited from the machine-tool floor — positioning accuracy and transfer speed — and worked its way into Matsushita (Panasonic), Sony and Toshiba.

By the end of the decade the shape of the company had changed without anyone announcing it. A Sendai office opened in 1986, an Okazaki plant in 1989, and in September 1990 Fuji moved up to the first section in Nagoya. Thirty years after founding it stood on two legs — machine tools and electronic assembly — with the weight already shifting to the second.

Through the 1990s the world market for placement machines grew at double-digit rates on mobile phones, personal computers and car electronics, and Fuji built out to meet it: a German subsidiary near Frankfurt in 1991, an American one in Illinois in 1994, a Brazilian one in 1995. The field was crowded — Panasonic, Yamaha Motor and JUKI at home, Siemens (later sold to ASM Assembly Systems) and Philips-owned Assembleon in Europe — and Fuji competed on modular construction, placement speed and the ability to handle ever-smaller components. By the late 1990s it stood among the leaders in units shipped worldwide, ahead of the machine-tool business that had created it.

The decision that carried it into the 2000s was to replace its own winner. The rotary CP series was losing ground to fixed-table high-speed machines, and in June 2003 Fuji completed NXT, a modular high-speed multi-function placement machine that could be reconfigured as a line’s product mix changed — the answer to shorter product lives and shorter runs. NXT carried the company’s share through the decade. Sales and service followed the assembly work east: Shanghai in 2007, Kunshan in 2012, and a Tokyo first-section listing in 2013.

The cost of that concentration showed in the crash. The Lehman shock cut electronics capital spending away, and in the year to March 2010 revenue nearly halved, from ¥69.5bn to ¥41.7bn, with a net loss of $55M (¥5bn) — the first consolidated loss since 1959. A year later revenue was ¥92.9bn and net profit ¥12.9bn. That whiplash is the business: mounter demand moves with the semiconductor and electronics investment cycle, and the whole company moves with it. Machine tools, meanwhile, held around a tenth of sales and posted segment losses again and again — ¥913m, ¥203m and ¥712m in the years to March 2011, 2015 and 2017 — while the assembly segment earned ¥24.8bn, ¥15.2bn and ¥14.5bn over the same three. In 2015 the segment was renamed Robot Solutions, signalling a move from selling machines to selling whole assembly lines, and in the year to March 2018 the group posted record revenue of ¥120.0bn and net profit of ¥17.5bn.

In August 2018 Fuji Machine Mfg. renamed itself FUJI and, in the same year, bought Fasford Technology of Minami-Alps, Yamanashi, taking it into semiconductor back-end assembly — die bonders, which join a chip to its substrate. President Soga Nobuyuki explained the renaming as closing the gap between a product brand that was well known abroad and a corporate name that still said machine-tool maker; his successor Suhara Shinsuke put it as leaving behind the image of a maker of production equipment altogether. The purchase was the argument the new name needed: precision positioning and machine vision, moved one step along the electronics chain into a market with a different demand curve.

Under Suhara, who became president in May 2019, the company set itself “three zeros” — zero placement defects, zero operators, zero machine stoppages — a target that only makes sense if you are selling an unattended line rather than a machine. FUJI Linear was set up in 2020 to develop linear-motor-driven placement machines, an Indian subsidiary opened in 2019 and the Singapore sales company was taken fully in-house in 2021. In April 2022 the shares moved to the Tokyo Prime and Nagoya Premier markets, and that June Soga returned as chairman and president, promising to make the machine tools general-purpose enough to sell.

Isozumi Joji became the seventh president in May 2023, with an ambition stated as customer behaviour: to be the first company anyone calls when they decide to build an electronic board. His Medium-Term Plan 2026 attached numbers to it — ROE of 10% and a price-book ratio above 1.1 by March 2027, a dividend floor of ¥80 and a payout ratio of at least 50% — aimed squarely at a balance sheet so cash-rich and debt-free that the shares traded below book. Revenue fell from a peak of ¥148.1bn in the year to March 2022 to ¥127.1bn two years later; the plan was kept. Alongside it FUJI began working with Silicon Valley startups on businesses that do not move with the mounter cycle at all — the Hug transfer-support robot, smart lockers — small so far, and intended as ballast.