Collaborating for Open Source Innovation

Article by Cindy Chepanoske, Jacob Green, Joe Doyle, John Whelan, Lucie Smolka, Clare Dillon, Colleen Maloney, Ciara Flanagan

At the recent OSPO++ ‘Open Source Innovation in Universities event’, Cindy Chepanoske (Director of Technology Licensing at the Center for Technology Transfer and Enterprise Creation, Carnegie Mellon University); Jacob Green, (Founder of OSPO++); Joe Doyle (IP Manager at Enterprise Ireland); John Whelan (ICT Research Commercialization Manager, Open Source Program Office, Trinity College Dublin); and Lucie Smolka (CEO, Open Cities) discussed collaboration in the context of open source innovation. The panel explored common themes emerging from their own organizations’ experience of collaboration at an institutional level either between universities, within academic-industry partnerships and even amongst cities.

Addressing the challenges of joint Intellectual Property

The challenges of collaboration between universities was acknowledged during the discussion. However, Cindy Chepanoske provided an example of a successful artificial intelligence project that involved international collaboration between students. Initially, the technology transfer offices (TTOs) of their respective universities started looking at intellectual property (IP) and copyright. However, Chepanoske suggested that everyone publish under an open source license.

“This was in the spirit of students who just wanted to work together … From the contractual side, it was solved very easily by having everything under an open source license.”

John Whelan concurred with this approach when discussing TerminusDB, a start-up that was originally funded through Horizon Europe. The Knowledge Transfer Office at TCD successfully obtained permission from all parties to allow the project to spin-out as an open source based company using the CORE business model. “Joint IP is murder and open source gets over the problems of joint IP.”

Whelan also referred to the importance of Knowledge Transfer Ireland’s catalog of template agreements. Trinity College Dublin regularly signposts funders to these agreements. While commercial companies collaborating with universities are not legally bound to use these agreements, the catalog is representative of Government policy and good practice in this area.

A template for open source licenses has not been developed by Knowledge Transfer Ireland and there was a call to action for an open source, intra-University collaboration to develop an open source model agreement.

As IP manager with Enterprise Ireland, Joe Doyle explained that too often he is contacted to help a company address an IP issue after the fact, “[We’re called on] to try and put the toothpaste back into the tube.”

Doyle outlined some common challenges for companies coming out of development programs that want to commercialize. In a number of cases, new companies may not understand the implications of the open source licenses they’ve already signed up to. Other companies may discover during the due diligence process that they don’t have the Manifest for open source code in their product.

Doyle stressed the importance of developing an IP strategy that incorporates an open source component from the outset. This may be as straightforward as developing a database of relevant open source licenses and their terms.

Balancing expectations and needs

A key issue discussed by the panel focused on the need to manage expectations in relation to the research. At times, funders and sponsors (particularly those from the commercial sector) seek assurances around the usability, rights and ownership of the research output at very early stages of a project.

Providing concrete assurances early in a project tends to be counterproductive. The research process itself involves exploration, discovery, testing and trialing of multiple approaches. Consequently, research may be hindered if researchers are limited by guaranteeing the code or models that will be used.

Chepanoske explained that researchers or universities are often unable to provide the kinds of assurances sought. “Sometimes, we’re asked what types of open source we’re building from. [The researchers] may not know at the beginning because it is research. Part of the research is exploring all of the models that are out there,” she said. “It’s important to have the freedom for researchers to move in a way that works for them.”

In these cases, robust conversations are required at an early stage to understand both the needs and expectations of funders or collaborators whilst also ensuring that researchers can undertake their work in an optimum research environment.

Creating sustainability through strategic collaboration

Lucie Smolka discussed Open Cities’ journey as an NGO promoting the creation, reuse and sharing of open source software and open data across municipalities.

There are clear parallels between this initiative and open source in the ‘real world’ environment. Smolka referred to approximately 100 projects that were originally set up by enthusiastic developers. However, in the absence of funding for ongoing support and maintenance, the projects could not continue.

Open Cities prioritized three projects and broadened its remit to partner with organizations. It embarked on a collaboration with Česko.Digital, a civic tech organization of expert volunteers who donate their time to supporting the state and the not for profit sector. These volunteers provide the support needed to maintain priority projects.

Open Cities is also collaborating with the Ministry of the Interior of the Czech Republic on the development of an open source portal for discovering and sharing information about relevant open source projects.

Smolka noted the importance of easing the pathway to collaboration with potential partners. “We really focused on making the projects as professional as possible. Everything is documented nicely to bring people on board. We’re creating these small communities around our projects with volunteers and we support them to continue the projects.”

The panel agreed that a strong community is key to sustainability.

“If you can create a strong community, it’s even better for open source than the commercial software,” Smolka explained. “But you have to focus on creating that community because it’s not easy.”

Community building and management in open source requires resources and can be time-intensive. However, it remains a crucial route to ensuring the future of a project, regardless of its commercial use.

The Open Source Program Office and collaboration

Jacob Green outlined the unique role that Open Source Program Offices (OSPOs) play in relation to collaboration in the university context.

OSPOs act as “Organizational APIs”, that build the capacity of faculty and students to “do open source,” through a range of activities including the provision of information, guidance and support.

As institutional constructs, OSPOs also enable collaboration through the formation of networks and/or by actively facilitating partnership working. In these cases, the collaboration itself is a tangible output - not only in terms of shared learning, resources and the benefits of being in a network but also in relation to the projects and code generated through these collaborations.

All agreed that as open source and open science gain traction and increasingly become the default, collaboration will continue to underpin successful OSS innovation, creation and maintenance.

View the complete OSPO++ panel discussion here.

This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0)