Down the White Rabbit hole of CERN’s precision time tech

This paper received several awards:  

• Best Paper 2023, AIS Special Interest Group Grounded Theory Methodology  
• Trailblazing Research Award 2023, This IS Research  
• Best Paper 2023, MIS Quarterly  
• Best Paper 2023, Association for Information Systems 

Open-source hardware (OSH) is revealing itself to be an increasingly promising phenomenon. However, limited insight into the use of hardware made freely available to study, modify, distribute, make and sell has hindered its commercial potential.  

To better understand how to overcome these shortcomings and boost open-source development, Esade’s Assistant Professor Laia Pujol Priego — who sadly passed away in 2023 — and Professor Jonathan Wareham analyzed the White Rabbit OSH, initially developed at CERN to eliminate distortions in time measurement across particle accelerators and computing networks in different geographical locations. The technology is now widely used in commercial industrial settings including automated vehicles, air traffic navigation and telecommunications networks.  

The analysis of White Rabbit’s evolution, published in MIS Quarterly, is the result of unprecedented access to the CERN Hardware and Timing Section at the Beams Department, which coordinated the development of the OSH.  

Open and hybrid

Over the last two decades, OSH has become increasingly influential in commercial hardware development. As examples, the Italian company Arduino, formed in 2005 as a low-cost way for students and hobbyists to access devices, now offers accessible electronics and digital technologies in partnership with global giants including Google, intel and Bosch. Or the Open Compute Project, which began as an internal Facebook project in 2009 and became a standalone foundation in 2011, is now a global datacenter IT infrastructure community.  

However, the development of physical open-source technology is not without problems. When such objects are developed with the use of middleware (to facilitate integration between systems) or embedded software (to control specific hardware) they result in the creation of hybrids. Pujol and Wareham argue that hybrids have a unique nature that introduces new challenges including long development cycles, high production costs and unclear licensing rules.  

The diverse community of technical specialists, commercial vendors, industrial complementors and voluntary contributions that make up White Rabbit offered the researchers a decade’s worth of rich longitudinal data in which to explore these issues. 

Following the white rabbit 

A collaborative approach to the development of White Rabbit was adopted by CERN from the outset. The project was launched as an open-source hardware initiative, accompanied by an open call to CERN’s vendor network to encourage contributions. To facilitate this collaboration, CERN developed an open-source hardware license, which established clear rules for sharing, distributing, and selling WR designs. A robust support framework was also implemented, including a centralized repository for documentation and design files, a Wiki for guidance, a developers’ mailing list for communication, regular workshops for knowledge sharing, and a suite of collaborative tools to streamline contributions and modifications. 

The White Rabbit community was quickly joined by a large and diverse group of organizations. Within a decade, the number of contributors had surpassed CERN’s ability to track the reuses and adaptations made to White Rabbit.  

With their unprecedented level of access to the White Rabbit community and associated data, Pujol and Wareham were able to define this development into three distinct phases: hybrid formation (2008 to 2012), hybrid development in local contexts (2012 to 2015), and the development of hybrid derivative works (2015 to 2020). 

Phases of evolution  

During the hybrid formation phase, it quickly became evident that CERN needed to centralize and control development. This work was coordinated by and integrated at CERN. Contractual agreements governed all developmental activities and specified that all documentation must be shared in the repository and governed under an open-source license. 

The second phase began with the release of the first prototype, at which time users began reporting bugs. Fixes were incorporated into further designs and novel versions began to appear based on the unique requirements of participating institutions. With many new designs and an increasingly heterogenous community, the White Rabbit development was characterized as being more loosely centralized and highly voluntary.  

During phase three, a standardization process was developed to guarantee the stability of the technology and raise awareness of White Rabbit’s potential outside the scientific research community. During this phase, industrial implementors gained greater autonomy over the modification of legacy systems and new versions were developed as proprietary applications that were not always disclosed to the White Rabbit community. 

Proposing a future for hybrids in OSH 

Following their analysis, the researchers developed a series of propositions to offer a more refined understanding of the role of hybrid objects in open-source development. The central insight from the analysis is the concept of malleability, which refers to the adaptability of a hybrid object to be adjusted, modified, or reconfigured while maintaining its core functionality and essential characteristics. Malleability is pivotal in understanding how hybrid objects evolve and accommodate diverse operational requirements within open-source development frameworks. 

The first proposition is that malleability it’s determined by three structural attributes: 

• Embodiment refers to the physical or digital nature of an object’s components. For example, physical components are subject to constraints such as size, cost, and material properties, while digital components offer greater flexibility for adaptation. 
• Modularity indicates how tightly or loosely the components of an object are coupled. Tightly coupled components are challenging to modify independently, while loosely coupled components can be adjusted or replaced without affecting the entire system. 
• Granularity is the degree to which an object can be decomposed into smaller, manageable components. High granularity means the object is composed of numerous smaller parts, which can be individually modified or upgraded, while low granularity indicates that the object functions as a single, monolithic unit. 

This characterization leads to the second proposition, that the malleability of a hybrid increases with maturity. In the case of Whit Rabbit, During the early stages of development, a high degree of centralization was necessary to retain cohesive structure and sequential steps. The research data clearly showed that a development model with multiple contributions would reduce operating precision and increase development efforts. But as the technology matures and is adopted in differing settings (such as high altitudes or deep-sea levels), its malleability continues to develop. 

The third proposition is that less malleable objects are associated with a development model characterized by assigned tasks and pecuniary compensation, centralization and sequential collaboration. Low malleability requires tightly controlled and centrally directed development processes as in the first phase of White Rabbit, as these objects necessitate sequential steps and coordinated efforts. 

In contrast, the fourth proposition indicates that the development of highly malleable objects is associated with self-assigned tasks, decentralization and asynchronous collaboration. This is consistent with the second proposition and the White Rabbit case, who suffer numerous independent modifications with maturity and higher malleability at more advanced stages of the project.  

Finally, the fifth proposition reflects the dual processes of liquification and crystallization that illustrate how these hybrids, once they have reached high malleability, evolve to balance open-source principles with the constraints of real-world industrial and legacy system integration. 

• Liquification describes a process where the core hybrid technology is stable enough to enable modifications of its peripheral components, allowing for changes that are generally consistent with the original open-source design. 
• Crystallization involves the hybrid technology adapting to constraints imposed by legacy systems or specific industrial standards. These modifications may result in tightly coupled, proprietary versions of the technology that deviate from the original open-source logical design. 

Challenging assumptions

This final insight of the authors stresses the importance of differentiating physical implementations of White Rabbit across its development from the design shared and regulated by an OSH license. The mutual influence of increasingly heterogeneous implementations and contributions to logical design has implications for third-party manufacturers and implementors, as well as further informing future versions of White Rabbit logical design.  

In this sense, the authors say, open-source hardware does not offer free digital, hybrid and material items — but rather provides the freedom to use and modify what is codified under the terms of the license.  

By observing the trajectory of OSH malleability, the research challenges previously held assumptions and offers a more nuanced explanation of how physical objects interact with other attributes. The insights, the authors conclude, offer new avenues to leverage the potential of OSH in both digital and hybrid objects and create a basis for a better understanding of development practices.