Sharing hardware designs is much older than the phrase "open hardware." Radio amateurs traded circuit diagrams in magazines for most of the twentieth century, early personal computer clubs passed around schematics, and computer manuals once routinely included full circuit diagrams. What changed in the 1990s was the arrival of free and open source software, which gave people a vocabulary and a legal model to borrow. This history follows that borrowing, with the dates that most accounts agree on. Where sources differ, we say so.
Late 1990s: certification and definitions
The first organized efforts weren't about publishing whole designs. Free software developers needed documentation to write drivers for hardware, and many manufacturers kept their programming interfaces secret. Around 1997 and 1998 (accounts give both years), a volunteer-run Open Hardware Certification Program offered manufacturers a way to self-certify that they would publish enough information for anyone to write software for their devices. Products that qualified could carry a logo. The idea was a promise about documentation, not about opening the circuit design.
That program's approach of self-certification, a public list of products and a logo for compliant ones resurfaced almost two decades later in OSHWA's certification, though the two programs were unrelated and asked for very different things.
Other projects in those years went further and aimed at sharing designs themselves. Open Design Circuits, launched in 1998, was a website for collaborative circuit design. The Open Hardware Specification Project, announced the same year, tried to define open hardware specifications. In 1999 a group of academics set up the Open Design Foundation and began drafting an Open Design Definition. Around the same time, OpenCores started as a community for sharing open digital logic designs, the IP cores used in FPGAs and chips. Several individuals published draft definitions of open source hardware in 1998 and 1999. None became a lasting standard, but they established the main questions: what counts as the "source" of a physical thing, and what freedoms should come with it.
2005 to 2009: cheap tools and big projects
Interest picked up in the mid-2000s, pushed along by cheaper PCB fabrication, inexpensive microcontrollers, and the spread of hobbyist tools. Two projects from 2005 became reference points.
Arduino began in Italy as a simple, low-cost microcontroller board for students and designers. Its board designs were published openly, and its friendly software environment made it popular far outside engineering. It also showed a business model that's still common: open designs combined with a protected brand name.
RepRap, started in 2005 at the University of Bath, set out to build a 3D printer that could print many of its own parts. Its designs were open from the start. RepRap and its descendants seeded a large part of the desktop 3D printing industry.
Licensing caught up during these years. In 2007, TAPR, a nonprofit amateur radio organization founded in 1982, published the TAPR Open Hardware License, generally described as the first license written specifically for hardware. At CERN, engineers began an Open Hardware Repository that went online in January 2009, and in July 2009 participants at a workshop in Banff, Canada, created OHANDA, the Open Hardware and Design Alliance, which offered a registration label for open designs.
2010 to 2012: a shared definition
The definition most people use today came out of a burst of activity in 2010. A small group of practitioners circulated a draft Open Source Hardware Definition early that year, workshops followed in New York, and version 0.3 was published in July 2010. The first Open Hardware Summit took place in New York in September 2010 with around 320 attendees. On February 10, 2011, the community released Open Source Hardware Definition 1.0, modeled closely on the Open Source Definition for software. A community logo, the gear-shaped "open source hardware" mark, was chosen by public vote that spring.
CERN released the first CERN Open Hardware Licence in July 2011. It was one of the first hardware licenses backed by a large institution, and it quickly found users beyond physics labs.
In June 2012, the Open Source Hardware Association (OSHWA) was founded in the United States as a nonprofit to host the Summit and look after the definition. The community didn't fully agree on every point during these years, and a separate group kept working under the "open hardware" name, but the definition and OSHWA became the most widely cited reference.
2013 to today: certification, licenses and institutions
OSHWA opened public discussion of a certification program in 2015 and launched it in October 2016 at that year's Summit. Projects that meet the definition can register for free, receive a unique ID and use a certification logo. The directory now lists projects from many countries. Our certification guide explains how it works.
Licenses matured too. The Solderpad Hardware License, a hardware adaptation of the Apache License 2.0, became a common choice for open chip designs, and its stewardship moved to the FOSSi Foundation in 2020. The same year, CERN published version 2 of its license in three variants, strongly reciprocal, weakly reciprocal and permissive, and all three were approved by the Open Source Initiative. The license guide compares them.
Meanwhile open hardware spread into new fields. Open instruments became a regular sight in research labs, where they cut costs and make experiments easier to reproduce. Open processor designs, helped along by open instruction set architectures, moved open hardware into chips. Educational boards and kits brought schematics into classrooms, and repair advocates pointed to open designs as proof that devices can be documented without sinking a business.
Patterns worth noticing
A few things repeat across this history.
Certification ideas keep returning because a label solves a real problem: "open" on a product box is easy to claim and hard to check. The late-1990s program and OSHWA's program addressed it in different ways for different eras.
Definitions came after practice. People shared designs for decades before anyone wrote down what open hardware meant, and the 2011 definition codified habits that projects like Arduino and RepRap already had.
Licenses took the longest. Hardware sits awkwardly between copyright, patents and trademarks, and it took several attempts, from the TAPR OHL in 2007 to CERN-OHL v2 in 2020, to produce licenses that lawyers and engineers were both comfortable with.
Where to read more
OSHWA keeps a detailed history of open source hardware organizations and definitions, which was one of our sources for the dates above. For what the definition requires today, start with our guide to what open-source hardware is.