The Unlikely Origin Story of FDM and SLA: From Garage Inventions to Global Manufacturing

The air in the small garage in Irvine, California, reeked of uncured acrylic and burnt solder on the afternoon of March 8, 1983, when Chuck Hull first ran a hand-held ultraviolet (UV) lamp over a drop of photosensitive resin he had smeared on the tines of a plastic spoon. What he saw next would upend 10,000 years of manufacturing orthodoxy: the resin hardened instantly where the light hit, leaving a thin, rigid film on the spoon’s surface while the rest remained liquid. Hull, a 35-year-old engineer working for a small coating materials firm, had spent months experimenting with UV-curable resins for furniture finishes, but this accidental experiment sparked an idea so radical he spent the next 11 hours scribbling calculations on the garage’s concrete floor: if he could control exactly where the UV light hit the resin, he could build a three-dimensional object one solid layer at a time, rather than cutting, carving, or casting it from a larger block of material.^[1]

For the next year, Hull worked 18-hour days in that garage, building a prototype machine from surplus lab equipment, old scanner mirrors, and a modified x-y plotter he scavenged from a defunct drafting firm. There were no commercial CAD programs to generate 3D models in 1983, so Hull hand-calculated the coordinates for each layer of his first test part—a small plastic cup—using a TI-59 programmable calculator, then punched the coordinates into the plotter’s controller one line at a time. When the machine finished its first successful print 12 hours later, Hull knew he had invented something that would redefine manufacturing, even if the rest of the world had no idea what to make of it. He filed the foundational patent for stereolithography (SLA) on March 8, 1984, and the core claim of that patent—*“a method of forming a three-dimensional object by successively solidifying layers of a fluid medium”*—remains the legal and technical bedrock of all additive manufacturing processes today, including fused deposition modeling (FDM).

The first commercial SLA system, the SLA-1, launched in 1986 when Hull founded 3D Systems with $1.5 million in seed funding. It was a behemoth: 1.2 tons of steel, glass, and electronics, with a 100mm x 100mm x 100mm build volume, a 100W mercury-vapor UV laser, and a price tag of $500,000—equivalent to $1.4 million in

The first commercial SLA system, the SLA-1, launched in 1986 when Hull founded 3D Systems with $1.5 million in seed funding. It was a behemoth: 1.2 tons of steel, glass, and electronics, with a 100mm × 100mm × 100mm build volume, a 100W mercury-vapor UV laser, and a price tag of $500,000—equivalent to $1.4 million in 2024 dollars. Its first customers were not hobbyists, but the aerospace and automotive giants that could afford to pay for a technology that cut prototype development time from weeks to hours. Ford Motor Company, for example, used the SLA-1 to print scale models of new car interiors in 6 hours, a process that previously required 4 weeks of hand-milling clay. Lockheed Martin used early SLA systems to prototype parts for the F-22 Raptor fighter jet, reducing the time to test fit for complex ducting components from 6 weeks to 2 days. These early wins proved SLA’s value for high-stakes, time-sensitive prototyping, but the technology remained out of reach for all but the largest corporations: the SLA-1 required a dedicated clean room, specialized ventilation for toxic uncured resin, and a full-time technician to operate it. By 1990, fewer than 200 SLA systems had been sold worldwide, and the technology was widely dismissed as a niche tool for only the most well-funded engineering teams.

While Hull was building the first commercial SLA systems, a parallel invention was brewing 3,000 miles away in the suburbs of Austin, Texas. In 1988, Scott Crump, a mechanical engineer and amateur inventor, was trying to build a custom toy frog for his 11-year-old daughter using a hot-glue gun and a modified wax dispenser. When he tried to melt a stick of polyethylene thermoplastic in the dispenser and extrude it layer by layer into a 3D shape, he realized he had stumbled on a far more practical alternative to SLA: a process that used cheap, non-toxic, off-the-shelf thermoplastic pellets instead of expensive, messy, and volatile photosensitive resin. Crump and his wife Lisa, a materials scientist, spent the next year refining the process, adding a heated build chamber to reduce warping and a computer-controlled x-y gantry to align each deposited layer. They filed a patent for *fused deposition modeling* (FDM) in 1989, the same year they founded Stratasys with $50,000 in seed funding. Unlike Hull’s garage invention, which was built to solve a problem in industrial prototyping, Crump’s FDM was born from a father’s desire to make a toy for his kid—a detail that would later become a core part of the technology’s identity as a tool for democratized manufacturing.

For the first 20 years of its existence, FDM lived in SLA’s shadow. Stratasys’ first commercial FDM system, the 3D Modeler, launched in 1991 with a price tag of $200,000, and it was widely panned by industrial users for its poor surface finish and low dimensional accuracy. Early FDM parts had visible layer lines as deep as 0.5mm, and the thermoplastic materials available at the time (mostly ABS and polycarbonate) warped badly as they cooled, leading to print failure rates as high as 70% for complex parts. Industrial engineers dismissed FDM as a “toy” for making low-fidelity prototypes, while SLA remained the gold standard for high-detail visual models and functional prototypes for aerospace and medical devices. By 2000, Stratasys had sold fewer than 1,000 FDM systems worldwide, almost all to large manufacturing firms, and the technology seemed destined to remain a niche alternative to SLA for the foreseeable future.

The turning point for FDM came not from a corporate R&D lab, but from a university workshop in Bath, UK, where Adrian Bowyer, a senior lecturer in mechanical engineering, was trying to print a custom bracket for a laboratory machine in 2004. The local 3D printing service charged him £400 and took 2 weeks to deliver the part, a cost and timeline that Bowyer found absurd for a simple plastic bracket. “I thought, there has to be a way to make these machines cheap enough that anyone can have one in their garage,” Bowyer later recalled in a 2019 interview with *Wired* magazine. He began tinkering with low-cost RepRap (short for *Replicating Rapid-prototyper*) designs, using off-the-shelf components from CD-ROM drives for motion control, PVC pipe for the frame, and Arduino microcontrollers (then a new open-source electronics platform) to run the machine. His core goal was not to build a better commercial printer, but to build a machine that could print most of its own components, creating a self-replicating system that could spread exponentially without centralized manufacturing.

Bowyer released the first full design files for the RepRap “Darwin” printer in 2009 under an open-source license, free for anyone to download, modify, and build. The first Darwin printer cost less than £500 in parts, a fraction of the cost of a commercial FDM system, and it could print roughly 60% of its own components, including the plastic frame parts, gears, and mounting brackets. The remaining 40% (stepper motors, electronics, and metal rods) could be sourced from online hobbyist suppliers for less than $200. Overnight, the barrier to entry for 3D printing collapsed. Within 6 months of the Darwin’s release, more than 1,000 Darwin printers had been built by hobbyists around the world, and a global community of volunteer engineers began iterating on the design, fixing flaws, adding features, and sharing improvements for free on online forums. By 2012, the RepRap community had spawned dozens of commercial spin-offs, including Ultimaker and Prusa Research, which sold pre-assembled FDM printers for less than