A decade ago, the scientific community recognized that to move from open access to open science, there was a need for free unrestricted access to scientific knowledge. This meant valuing, sharing and preserving data, software and other digital artifacts from research, but the on-ramp to participate had to be faster and simpler if the practice was going to gain traction.
The European Union decided to fund CERN (the European Organisation for Nuclear Research) through the OpenAIRE project to build a catch-all repository to ensure all researchers had a place to easily upload software, data, preprints and other digital outputs.
That was the beginning of Zenodo, which CERN and OpenAIRE launched in 2013. Since, the free global platform has expanded faster than imagined. It now has 25 million visits a year, hosts 3+ million uploads and over 1 petabyte of data. This year marks the platform’s 10th anniversary and today Zenodo is widely viewed as a trusted place to preserve research materials that could be of use to others in advancing science.
At CERN, Lars Holm Nielsen, software engineer and project manager, and Tim Smith, project executive and group leader, were responsible for realizing the new repository. Nielsen had just left a position at the European Southern Observatory (ESO) where he had worked with Christopher Erdmann, a librarian, on helping to make astronomy data open and discoverable for both the public and scientists. They stayed in touch after they both left ESO.
Although an ocean apart, the three were instrumental in the creation of Zenodo — Nielsen at CERN did the technical development, while Smith and Erdmann designed solutions for challenges and focused on community outreach. Both OpenAIRE’s base in Europe as well as Erdmann’s base in the U.S. and large professional network were critical in gaining traction for Zenodo.
After considering several more generic names (Research Share, for example), the system was named Zenodo, a nod to Zenodotus, the first librarian at the Ancient Library of Alexandria and first recorded creator of metadata.
The repository was promoted as a one-stop-shop that welcomed research from all over the world and from every discipline. It accelerates the sharing of science by allowing researchers to share artifacts anytime without having to wait until publication of results.
“It’s a platform built to empower users to get things done, preserve research data and support science,” Nielsen said. “We wanted to make sure there wasn’t any excuse not to share data.”
By assigning each item a Digital Object Identifier (DOI), Zenodo provides a valued service of curating software, data, articles, conference materials—anything needed to understand the scholarly process. Scientists are able to get credit for important steps in the process they contributed to rather than the current ‘everyone or no-one’ on the paper author-list approach. Uploads are made available online immediately and the DOI is registered within seconds.
“Researchers were having a hard time going through all the steps to get data into a repository,” said Erdmann, now living in North Carolina and associate director of open science for the Michael J. Fox Foundation. “We wanted to provide a more flexible approach and to remove some of the barriers so people would just start sharing.”
In the past decade, Nielsen, Smith, and Erdmann continued to collaborate on Zenodo, working in particular with Jose Benito Gonzalez Lopez and Alex Ioannidis in a small team that helped expand Zenodo in response to the exponentially growing demand. The system was developed under the European OpenAIRE program and is operated by CERN for the research community worldwide, relying on CERN’s reputation and large-scale data management expertise, adding to Zenodo’s credibility and stability.
Initially, Zenodo was seen as a disruptor of sorts.
“Letting researchers curate themselves was a radical thought,” said Smith of their concept. “This was a change in the process, relinquishing control and lowering the barriers for people to submit.”
Yet, participation in Zenodo exceeded the expectations of many—especially early critics who thought it was too open. Over time, it has won over skeptics as researchers told others and the ease of use drew more to deposit their work.
“The ‘communities’ feature in Zenodo allowed anyone to create a repository for their resources and it helped lower barriers for the greater community to start sharing research,” Erdmann said.
Scholars began to share materials from meetings and conferences on Zenodo with one link so everything was discoverable and citable.
Among the examples of impactful scholarship shared on Zenodo:
a 3D-printable model of the universe, a COVID literature review study and software used as part of the Event Horizon Telescope.
Zenodo’s technical platform was open source from the beginning, but was never intended to be used by others. Yet, institutions started using it to build their own repositories because they wanted to provide the Zenodo-experience to their own researchers.
That sparked Gonzalez Lopez and Nielsen to get a small grant from CERN Knowledge Transfer in 2018 to make the platform easily reusable. Today, the platform named InvenioRDM is co-developed by a collaboration of 25 research institutions world-wide–a win-win for both Zenodo and each of the institutions.
“It’s important that we have an open, community-run infrastructure,” Gonzalez Lopez said. “Our mindset was for the repository to be used all over the world by people who don’t have resources and level the playing field.”
Nielsen said he hopes Zenodo continues to be a useful place for researchers to share their work and that the platform continues long after his involvement.
“Problems come from getting bigger,” Nielsen said. “My vision for Zenodo is to keep solving those problems while we stay innovative at the leading edge of scholarly communication.”