It took about 5,000 years from the discovery of glass until a process was developed to economically mass produce it in large sheets. It took barely a decade before the technology was stolen.

One of the great fundamental inventions–not at the level of the wheel or fire, but pretty high on the list–glass is chiefly made from relatively common and inexpensive raw materials: sand, soda ash, and lime.

It’s hard to think about glass as so valuable that someone would go to great lengths to steal the technology to produce it. But that’s what happened–twice.

First, a U-M alum staked his company’s future on violating a patent and won. An Ann Arbor engineer who twice tried to sell the technology to other countries had worse luck. After an FBI investigation, he was caught and prosecuted–twice.

Glass has been used to glaze ceramic vessels since at least 3,000 B.C. By 1500 B.C., Egyptians produced glass vessels (which in time they would use to hold beer, one of my favorite inventions). Glassblowing was discovered by 30 B.C.

But it took nearly 2,000 years to develop a way to economically create large sheets of flat glass. In 1848, Henry Bessemer, an English engineer, designed a system that formed a ribbon of molten glass between rollers. It worked, but the finished glass was expensive because it had to be ground and polished before use.

The real breakthrough came in the late 1950s, when Sir Alastair Pilkington and Kenneth Bickerstaff of Pilkington Brothers, Ltd. in the U.K. developed the first successful commercial “float” glass process.

The Pilkington process heats a mix of raw materials to about 1,500 degrees Celsius. Once it melts, it is allowed to flow onto a “tin bath” that is typically three or four meters wide and fifty meters long. Literally floating on top of a layer of molten tin in an oxygen-free atmosphere, it spreads out to a uniform thickness.

As the glass flows along the tin bath, the temperature is gradually reduced. At the end of the bath, it is solid enough to be lifted from the bath onto rollers. It then passes through a lehr (a type of kiln), where it gradually cools, and is cut by machines.

Pilkington Bros. filed many patents to recoup the substantial cost of development. Today, virtually all flat glass is made using the float glass process, including the glass for car windshields and windows. That’s where Michigan enters the story.

Since the time of Henry Ford, Ford Motor Co. had embraced the industrial strategy of vertical integration. Ford didn’t just build cars, it produced as many of the car’s components as possible–including the glass and fabrication of the windshields and windows.

Ford quickly realized that the float glass process was not only more efficient, it produced superior glass with consistently better visibility–a very important characteristic for windshields. Ford licensed the process from Pilkington and built a float glass plant at its Rouge complex.

The new, more efficient process was a boon to Ford, but a threat to Guardian Glass Co., a small, Michigan-based fabricator of windshields. Founded in 1932, it had been put into bankruptcy in 1957 by a new CEO, Bill Davidson–a U-M grad and nephew of the company’s owner. Guardian emerged from bankruptcy in 1960 but continued to struggle.

Although the company had obtained the contract for all the windshields for American Motors, it was buying the glass it used from glass producers, primarily Ford. Davidson concluded that to be more profitable, Guardian would have to produce its own glass.

Davidson set out on an audacious plan. He hired away Ford’s chief glass engineer, Ed Sczesny, and launched plans to build a float glass plant in Carlton. Clearly, Sczesny brought proprietary trade secrets with him.

Guardian made a perfunctory request of Pilkington to obtain licensing rights, which Pilkington denied. Pilkington threatened to sue if Guardian moved forward with the float plant, but Davidson gambled that it wouldn’t follow through. If Pilkington sued and lost, it risked jeopardizing its entire worldwide licensing program.

Construction on the float plant began in 1969 and was completed at a cost of about $17 million. By the middle of 1970, the Carlton facility was producing 350 tons of glass per day. The following year, Guardian reached an ongoing royalty licensing agreement with Pilkington, but at substantially lower amounts than the other U.S. licensees paid, on the grounds that Pilkington had not provided any technical assistance to Guardian. (This must have been troubling to Ford, which was paying the higher licensing fees and also had lost its chief glass engineer.)

The gamble paid off. Guardian, since 2017 a division of Koch Industries, is today one of the four largest manufacturers of flat glass in the world, with 18,000 employees and plants in twenty-one countries.

Davidson became a billionaire. He owned the Detroit Pistons during their heyday as NBA champions (those teams were known as the “Bad Boys”) and before his death in 2009 donated millions of dollars to various causes and institutions, including $55 million to the U-M. The nonprofit William Davidson Institute at the business school bears his name, as does the athletic department’s player development center, housed in a cylindrical glass-walled showcase next to Crisler Arena.

Ann Arbor resident John Akfirat was aware of what Davidson and Sczesny had done: he, too, was a glass engineer at Ford. In 1973, he stole the blueprints for the Ford float glass plant and went looking for a buyer.

Ford discovered Akfirat was trying to sell its plans and his expertise and notified the FBI. The FBI used an undercover agent posing as a potential buyer to approach Akfirat. The plans were seized, and Akfirat was arrested and prosecuted. He pleaded guilty and served a short jail sentence.

In the 1980s, a foreign source reported that a float glass plant was being built in Romania. (This was during the Cold War, when Romania was part of the Soviet Bloc.) The source also reported that Akfirat was involved in the construction of the plant. Reportedly Akfirat had been traveling to Romania during the period when the plant was being constructed, 1975-1981. The case was presented to the U.S. Attorney for the Eastern District of Michigan, who authorized an investigation.

Primarily based on the source information and Akfirat’s past record, a search warrant was obtained for Akfirat’s residence off Geddes Rd. Agents found plans for a float-glass plant and documents indicating Akfirat had been in contact with the Romanians and had travelled to Romania to help with the construction of the glass plant.

As the case agent on the investigation, I took the plans to Pilkington Bros. in St. Helens, England (near Liverpool). Pilkington engineers told me that although they were for a plant smaller than any built by Pilkington or its licensees, several unique characteristics were based on the company’s patented and/or secret technology.

Akfirat was again prosecuted and again pleaded guilty in U.S. District Court to one count of wire fraud. He was cooperative, and maybe a mitigating factor was that the Romanian plant was never successful in producing quality flat glass, or maybe the judge understood that Akfirat was just trying to execute a double steal–the baseball strategy in which two base runners attempt a stolen base on the same play. He was sentenced to probation.

Akfirat was a naturalized U.S. citizen of Turkish origin. According to a 1984 Observer article, he left for Turkey after his sentence was completed.

Akfirat had been charged federally with wire fraud because at the time there were no federal statutes specifically addressing the theft of trade secrets.

That changed in 1996, when Congress passed laws criminalizing trade-secret theft and other forms of economic espionage. I’m not aware that criminal prosecution was ever contemplated for Davidson and Sczesny’s theft of trade secrets, but had these laws been on the books, they could have been prosecuted.

Greg Stejskal retired as special agent in charge of the Ann Arbor FBI office in 2006. A version of this article appears in his book FBI Case Files Michigan, to be published in June by the History Press.