Combining development, capacity building and responsible innovation in GCRF‐funded medical technology research

Abstract Development‐oriented funding schemes such as the UK Global Challenges Research Fund (GCRF) have opened up opportunities for collaborations between low‐middle income countries (LMICs) and high‐income country (HIC) researchers. In particular, funding for medical technology research has seen a rise in previously under‐represented disciplines such as physics and engineering. These collaborations have considerable potential to advance healthcare in LMICs, yet can pose challenges experienced to researchers undertaking these collaborations. Key challenges include a lack of tradition of HIC/LMIC collaborations within participating departments, lack of experience with development agendas, designing contextually‐appropriate technologies and ensuring long‐term viability of research outputs. This paper reflects on these key challenges, using the experiences of the authors on the Open Laboratory Instrumentation (OLI) project as a focalizing lens. This project was a GCRF‐funded collaboration between physicists in the UK and engineers in Tanzania to develop an open‐source, 3D‐printed, fully‐automated laboratory microscope. The paper highlights key ethics lessons learnt.


| Ethics and digital diagnostics
The advances in healthcare technologies has occurred against the backdrop of the "4 th Industrial revolution (4 th IR)." The World Economic Forum describes the 4 th IR as "the advent of "cyberphysical systems" involving entirely new capabilities for people and machines." 1 The 4 th IR represents entirely new ways in which technology becomes embedded within societies and even our human bodies. The 4 th IR raises a wide range of ethics concerns relating to the design and distribution of technologies. 2 In response to these challenges, there is growing global recognition of the need for governance and for educating researchers and innovators to responsibly produce technology that benefits society. Highly influential frameworks, such as Responsible Research and Innovation (RRI) have been introduced.
RRI recognises that the technologies of the 4 th IR are changing the way we live and work. 3 RRI was defined by von Schomberg as: "a transparent, interactive process by which societal actors and innovators become mutually responsive to each other with a view on the (ethical) acceptability, sustainability and societal desirability of the innovation process and its marketable products (in order to allow a proper embedding of scientific and technological advances in our society)." 4 Within the RRI framework practices of societal co-production and anticipatory harm mitigation are key. Through these practices RRI aims to foster the "design of inclusive and sustainable research and innovation." 5 Key to the RRI approach are the stages of reflection. The model for these stages was initially proposed by Stilgoe, Owen and Macnaghten in 2013, and comprises four interlinking stages: anticipation, reflexivity, inclusion and responsiveness. 6 Later adaptations of the RRI model, such as that proposed by the EPSRC, adapted these stages to anticipation, reflection, engagement and acting ("AREA"). Successful implementation of the RRI requires considerable investment from all stakeholders. Not only to research teams need to engage in critically reflexive dialogue, but also foster two-way dialogue with non-academic stakeholders. This dialogue focuses on opening up the research visions, impacts and questioning to broader deliberation, dialogue, engagement and debate in an inclusive way. 7 Like many ethical frameworks, the RRI model is not prescriptive of how the AREA practices are implemented within a specific project.
Indeed, RRI continues to be a largely top-down approach guided by standardized principles. 8 Nonetheless, there are a growing number of case studies demonstrate the heterogeneity of the transition from model to practice. 9 To date, however, RRI has mainly been applied to high-technology innovation in the Global North. As mentioned by Hartley et al, "there is a significant gap in the literature relating to the application of RI in low technology, Global South contexts, 10 or outside Western liberal democratic contexts." 11,12 The focus on high-technology and high-income countries  The GCRF funding scheme received considerable support for its focus on "blue skies" research on key issues such as secure and resilient food systems, sustainable health and well-being, inclusive and equitable education, clean air/water/sanitation and sustainable energy. Its potential was also recognized as an important contributor to research equity, and a key means of diversifying the research undertaken as North-South collaborations. In particular, it facilitated the inclusion of research outside of the well-established medical and healthcare collaborations and networks to include other fields such as engineering research.
The expected outcome of the funding scheme was to facilitate the deployment of UK research excellence to address the most "significant and complex problems faced by the developing world, while at the same time strengthening research capability in developing countries." 25 In order to do this, it explicitly 16 This is a term used to describe the practice of researchers from high-income countries who collect specimens/data from low/middle-income countries yet conduct all the analysis and subsequent storage out of the country of origin.

| Mixing research and development agendas
The GCRF makes an important contribution to the establishment of a wide-ranging research agenda covering a range of development issues. Nonetheless, to be properly realised, GCRF-funded academics to demonstrate specific skills and experience of working effectively with colleagues and partners in the Global South. 26 For many academics, this represents a departure from their prior research experiences, and requires a shift from more traditional roles of academics as passive observers of development practice. Instead, they are increasingly expected to take an active role in contexts where networks of actors with varying levels of power and capacity are competing and collaborating with one another to bring about change for specific groups of people. 27 In order to be effectively supported in these activities, it is important that academics have access to training on how best to run development-focused activities or research projects.
Key skills required for robust development agendas include getting buy-in from local communities, working with both academic and non-academic partners, coordinating and managing relationships between delivery partners. Some of these challenges are detailed in Table 1 below. Successful development activities also require systems to monitor and evaluate the success of these development agendas. These requirements also reflect the recognized challenges of effective equitable research partnerships. Implementing frameworks for equitable research partnership are recognized to require expertise, time and resources to implement. Indeed, a range of papers have highlighted the critical importance of leadership, resolution and resourcing in the successful implementation of equitable partnership relations. 28 Interestingly, these challenges have been recognized by the UKRI since the GCRF inception. 29 Nonetheless, the responsibility for supporting individual academics has been largely devolved to individual institutions rather than as a centralized support from the funding body.
It is important to recognize that the GCRF scheme has also ex-

| OLI AS A CASE STUDY
Receiving GCRF funding for health technology development is thus more complicated than simply delivering excellence in research. The The lack of guidance on how to enact commitments towards equity, responsibility, development and openness is compounded by traits within ethics discourse. Indeed, discussions around these topics remain largely distinct and lack explicit inter-topic linkages. 30 This an mean that different frameworks are used in response to different ethical issues, which can be both confusing and frustrating. The lack of guidance can be the cause of considerable ethical stress for researchers receiving these funds, particularly researchers from disciplines that do not have long-standing traditions in engagement with (bio)medical ethics, capacity building, participatory research or LMIC/HIC collaboration. In order to unpack some of these ethical tensions we now turn to a case study project recently completed by some of the authors.
The Open Laboratory Instrumentation (OLI) is a recent GCRFfunded project that was a collaboration between many of the authors on this paper. This collaboration involved physicists in the UK and engineers in Tanzania In order to achieve the goals of microscope design and sustainable distributed manufacturing, OLI was designed as an Open Hardware (OH) project. OH evolved from the Open Science movement and commits to using open research practices to (re)design physical technologies. OH designs (ie. mechanical drawings, schematics, bills of materials, source code etc), in addition to the software that drives the hardware, are all released under free/libre terms. This means that the technology can be shared, adapted and constructed by downstream users at no additional cost.

| Ethics and the OLI
The innovative approach to microscopy design, the intention to establish manufacturing practices in a LMIC (Tanzania) and the commitment to OH meant that the OLI clearly met both RRI and GCRF objectives. Nonetheless, integrating these two key areas of ethical responsibility into the project was not without challenges. The researchers on the OLI project came from engineering and physics background and, as many GCRF recipients in these disciplines, had no prior experience conducting research with a development agenda.
Moreover, while the UKRI promotes frameworks mediating technology research such as RRI, access to training in these frameworks varied. The experiences from the OLI project thus enables reflection on a number of different issues. First, it provides the opportunity to reflect on the challenges experienced by researchers who are expected to adopt a development-focused research agenda without prior training in development activities. Second, it allows critical reflection on how RRI practices are implemented by researchers working on collaborative co-design projects between HICs and LMICs. These are discussed in more detail in the following sections.

| Addressing development commitments
The development agenda of the GCRF related to the OLI focus on digital diagnostic technology fabrication. In order to address this, the OLI project needed to demonstrate the feasibility of designing an OH, 3D-printed microscope that could be fabricated to the same specifications and standards in both the UK and Tanzania ("proof-of-concept").
Nonetheless, to truly achieve long-term sustainable development, the researchers needed to demonstrate that this fabrication could continue after the end of the GCRF funding (long-term sustainability). As the GCRF funding was capped at 3 years, these two expectations meant that the researchers needed to reconcile the joint expectations of producing a working microscope and fostering long-term development capacity building became topics of lengthy discussion.

| Proof-of-concept
The OLI project linked directly to the development agenda of the GCRF through the potential to build capacity for digital diagnostic technology manufacturing in Tanzania. In order to realise this goal, the project needed not only to demonstrate proof-of-concept (ie. that this was feasible within a Tanzanian context), but also to illustrate how sustainable distributed manufacturing could be taken forward after the end of the project. To manage, reflect on, and learn about the effects and impacts they are having (or not) on development processes in the Global South How to manage impacts without being trained in "stakeholder engagement", "adaptive management," or the "theory of change" In order to demonstrate proof-of-concept for distributed manufacturing capabilities the researchers set up reliable and sustainable prototyping/manufacturing systems in two very different contexts-the UK and Tanzania. 33 These activities foregrounded a range of challenges within the Tanzanian context that were less prominent in the UK context and were related to complicated supply chains for hardware acquisition. These included unreliable/delayed delivery, quality/cost compromises and high importing duties.
While issues of supply chains challenges were common to the were key to navigating communication throughout the project.

| Longevity of project
The careful balancing act between openness, equity and pragmatism became apparent when the OLI team planned for the longevity of the project. While the openness of the technology remained a priority for post-project plans, it was recognized that the long-term production of the microscope in Tanzania

| Ethical concerns associated with achieving development commitments
The development focus of the GCRF grant scheme was taken very seriously by all researchers on the OLI project. Nonetheless, without specific guidance on how to undertake effective development activities, the researchers had to rely largely on their own interpretation of how to put these commitments into practice. Their learning experiences in this realm highlighted three key areas that gave rise to continual reflection and concern.
3.6 | Understanding the context  hearsay. They could also have conducted poor PAR by 'extracting' views from research stakeholders, rather than engaging in dialogue.
Such observations necessarily raise considerable concerns for GCRF projects. Effective RRI requires community participation in order to consider context, anticipate impacts, reflect on purposes, engage inclusively, and act responsibly. How these communities-be they clinicians, educators or researchers-can be effectively engaged in highly technical projects with short timeframes is challengingparticularly so in countries with low science literacy and little tradition of citizen engagement with research.

| Addressing capacity building goals
One of the key objectives of the GCRF fund is to "strengthen capacity for research, innovation and knowledge exchange in the UK and developing countries through partnership with excellent UK research and researchers." 42 This placed an expectation on the OLI project that it would contribute to research capacity building within Tanzania and contribute to knowledge exchange. The OLI team interpreted this commitment in three key ways, first to facilitate learning opportunities for early career researchers on the project, second to foreground equity within the collaboration and third to facilitate knowledge exchange through Open Research (OR) practices. Integrating these commitments into the project design led to instances of ethical concern.
3.10 | Integrating Open Research practices into cross-cultural collaborations Discussions around OR are relatively well-established in the UK, and researchers have access to considerable support to assist the integration of OR practices into their projects. In contrast, OR topics tend to be relatively new for many LMIC researchers, and there is little in the way of institutional support to assist them to integrate OR practices into their work.
The OLI project team thus had to navigate a number of complexities relating to the integration of OR practices into their research design. OR was not only a requirement of the GCRF funding, it was also an integral part of the project as an OH activity. Nonetheless, how openness is integrated into research can vary considerably, and there is considerable scholarship describing OR practices as a continuum rather than a check list. 44 Recognizing this led to two key realisations, first that the researchers more familiar with OR should not simply impose "ways of doing" onto the rest of the collaboration team. Second, that there were many ways in which the same goals of openness could be achieved.
While open work practices remained an integral part of the OLI rollout, the enactment of openness became a site of dialogue and constant revision. It required a complicated process of negotiation and compromise that relied on dialogue, trust and empathy. As openness is a personal value, strategic research approach, a commercial strategy and a means of communicating with stakeholders, finding a middle-ground approach can be difficult. Finding acceptable means of operationalizing openness in research is as much a practice in self-discipline as in compromise.

| Achieving equity in project hierarchies
The overall goal of the GCRF funding was to support equitable, collaborative blue-skies research. Nonetheless, when submitting the OLI grant it became apparent that adding the Tanzanian colleagues as co-investigators was problematic on the required submission platform as they were not associated with a research institution. While the Tanzanian team were equal partners in terms of research decisions, this bureaucratic issue meant that they were positioned as "sub-contractors" in terms of the grant administration. Without significant effort from everyone involved in administering the OLI grant, this could have led to transparency issues in relation to budgeting and reporting. In addition, the structure of the project related to the bureaucracy involved in grant administration. This led to complications in making the payments to the Tanzanian colleagues, leaving the team with cash flow issues as they waited for the money to clear.

| Ethical challenges associated with capacity building commitments
One of the key objectives of the GCRF funding was to "strengthen and professional development. 48 Navigating such issues is often compounded by a lack of training for the researchers involved, and the absence of dedicated funds and support to assist in efforts.
During the OLI project the researchers had to deal with a number of these issues throughout the project. These manifested through the design of the grant bureaucracy, variations in project priorities and expected outcomes and infrastructural differences.
3.13 | Navigating value-heterogeneity within the project team Effective cross-cultural dialogue is an area of extensive study, as are effective cross-cultural/national collaborations. 50 This literature foregrounds the importance of personal openness and engagement in relationship building as a means of ensuring meaningful dialogue. It is recognized that this is not only a continuous process, but also a personal commitment. 51 This requires considerable support and investment to ensure that the inter-personal relationships are in place before the start of the grant, rather than negotiated solely during the research process.
It must be recognized that even in situations where researcher values appear to align, misunderstandings can still arise. In the OLI case study, the collective commitment to OR practices led to many discussions and compromises about how to operationalise openness within the project. While value-alignment challenges are starting to be discussed in relation to the GCRF, these tend to focus on differences between global priorities and local realities rather than differences in personal preferences. Grieve and Mitchell, for example, offer the example that "aspects of global priority such as human rights and what these encompass may conflict with the reality on the ground." 52 It would thus seem that navigating value differences within the collaboration team are crucial to effective RRI engagement. Nonetheless, effective communication skills training and conflict resolution, as key skills for maintaining value-focused discussions are rarely offered to researchers.
3.14 | Creating a working team and addressing power asymmetries All the members of the OLI team mentioned the challenges of conducting research in Tanzania. These included infrastructural issues such as the delivery of project materials and lack of reliable power and internet. As the OLI team had spent considerable time developing interpersonal relationships, they were able navigate these issues constructively. Nonetheless, these situations continued to present instances of personal and moral unease. All members of the OLI team agreed that the instigation of regular meetings was vital in navigating these issues. Without these meetings, as well as the situational empathy developed through on-site visits, it is possible that these resource asymmetries could have evolved into power asymmetries that impacted on the efficacy of the collaboration.
Appreciating the infrastructural inequality between collaborators in LMICs and HICs and successfully navigating these issues in a sensitive manner is rapidly becoming a "take home" message of the GCRF funding scheme. 54 While such issues have long been discussed in relation to research capacity development, 55 there was a marked lack of training for researchers funded on the scheme. Similarly, current RRI discussions tend to focus on power asymmetries between the research team and user-public, or issues of gender and inclusion. They do not necessarily focus on power asymmetries due to different working environments.
The experiences of the OLI team foreground the need for better training in relationship managementboth in terms of GCRF funding and RRI. There is a need to support lead researchers in team building  51 The UK team did participate in a training day on working across cultural divides, that they found was a great opportunity to reflect and improve their practice. Nonetheless, continuous support for these activities was lacking. activities that avoid imposition of values and assumptions. In situations of infrastructural asymmetry nothing can be taken for granted, and one should not assume that communication is optimal between team members.

| Differing narratives and priorities
The OLI project team represents an interesting challenge to current RRI discussions due to the differing priorities of the two national teams. While the UK team was primarily focused on research, the Tanzanian team was focused on the commercialization of the microscope. Thus, within the same project team research and innovation was variously prioritized amongst individuals. Luckily, these interests were not exclusionary, and the team were willing to engage with activities outside of their priority area.
Nonetheless, the existence of these differing priorities within the same project team foregrounds a challenge to RRI discussions. While it is tempting to think of differing end-point priorities being negotiated between different stakeholders (ie. researchers and commercial companies), there is less guidance on how these negotiations occur within research teams.
A further challenge to navigating differing priorities is the role that technology plays in the narratives of the GCRF and RRI. The GCRF takes a generally positive view of technology as an enabler of development. In contrast, RRI discussions tend to take a precautionary approach and focus on ameliorating the negative impacts of sensitive technologies. Both, it would seem, tend to overlook the unequal distribution of the benefits of innovation and more contextual rich understandings of the flow of information through societies.

| Addressing funding structures and underlying grant frameworks
While the GCRF funds are intended to support equitable collaborations between the UK and DAC countries, the funding scheme is located in the UK. The awarding, administration and reporting of these grants thus follows the UKRI model. This includes specific requirements regarding grant administration, finance and ethics approval that the DAC collaborators are expected to follow.
These expectations can place considerable strain on the research collaboration. As was evidenced by the discussions within the OLI team, these expectations were often difficult to implement. The Tanzanian collaborators, for example, had to rapidly adapt to the required methods of expense reporting, while the UK collaborators had to act as a mediator between the project team and their university finance department. In such cases, it needs to be recognized that these structures are imposed on the DAC researchers as a non-negotiable aspect of the grant, and may not always be the priorities of those accepting the grant. Indeed, for many LMIC academics these represent a heavy burden as they attempt to align these expectations with their institutional bureaucracy that may not be equipped to provide them with the support they need.
Navigating such situations requires considerable commitment by all researchers, and a willingness to adapt and support each other so as to find an effective way forward. It is important to note that this success, in part, relies on the dispositions of the individual researchers exemplified by virtues such as humility, openness and curiosity. Such virtues are difficult to quantify, and definitely not

| CONCLUDING COMMENTS
The rapid expansion of health technology development continues to revolutionize healthcare provision. This impact is both global and local, and offers considerable potential for LMIC healthcare to advance through the ability to connect, rationalise and enrich existing health systems. Moreover, the potential for in situ design of International funding, such as the GCRF scheme, have been hugely influential in supporting health innovations in, and for, LMICs.
These funding schemes, however, are not without challenges.
The experiences of the OLI project highlight important ethical challenges that range from the structure of the grants to the design of contextually-appropriate technologies. What is apparent from the discussion above is that these different areas of ethical concerns are currently covered by a range of different frameworks and/or fields of discourse. These include Open Science, development agendas, RRI and equitable partnership. While these different areas are, of course, linked by common values such as fairness, honesty and beneficence, the positioning and framing of these common values differs between these areas. Moreover, unified discussions that demonstrate how these areas are simultaneously enacted within a research context are lacking. This can leave researchers not only confused about how to respond to the differing commitments implicit in their grant, but worried that they are continually failing to live up to expectations. This realisation is particularly concerning when considering researchers who have no prior experience in, or training for, conducting research within these different frameworks.
Nonetheless, these challenges are not insurmountable. Indeed, the need to enrich ethical guidance around the development of LMIC-appropriate health technologies can-at least in part-be done by linking together current areas of ethics scholarship. In particular, bringing these different areas of scholarship together to discuss these issues would be of considerable importance. For instance, the OLI team difficulties experienced in identifying OR practices that could be implemented by scholars in different research contexts with differing levels of computer literacy. Such issues could be offset by linking scholars on OR with development/capacity building scholars wellversed in LMIC/HIC collaborative partnerships. Similarly, the difficulties of engaging the public in the OLI project could be offset by linking RRI scholars to ethicists specialising in community engagement and long-term public-academic relationships.
What becomes additionally apparent from the OLI project is the urgent need to integrate health technology research into current health research/ethics dialogues. As many researchers working on health technology projects come from disciplines not traditionally engaged in medical ethics discourse there is the potential that the wealth of expertise in this area will remain under-utilized. In particular, key topics such as power dynamics/hierarchies in collaboration, LMIC/HIC collaborations, longterm participant and community engagement are topics that would enrich and support health technology researchers.
Collaborations foreground issues of research environment asymmetry and resource allocation, infrastructural challenges and societal engagement that require careful consideration. This opens the potential for researchers in the HIC who have all the money, resources and taken-for-granted infrastructure to set the terms, expectations and ways of doing things and impose them on the LMIC team without discussion, perhaps buying into the western 'rescue' narrative towards LMICs and Africa in particular.
In addition to these challenges, the OLI project highlights the need for an enrichment of current RRI discourse so as to provide guidance for researchers designing technologies for use in LMIC contexts. Some of the issues raised by the OLI project, such as sustainability after funding, integration into LMIC socio-technical systems, and public engagement, foreground the need for more thought on applying RRI in highly disparate contexts. They also highlight the critical need for training around managing research versus innovation expectations-both within the project team and beyond. These are issues that need to be further addressed by ethics communiteis around the world.
Linking RRI discussions to the current innovation ecosystems within LMICs, as well as the considerable expertise of NGOs and development specialists would not only enrich discussion, but support researchers with these aspects of their projects. 59 Recognizing the enormity of this field will allow researchers not only a good idea of positive actions, but also a level of humility about what they are likely to achieve. As highlighted by Franzen et al, research capacity building outcomes need to be equally valued with research outcomes. 60 The experiences from the OLI project provide much scope for reflection on research funding schemes for medical technologies that include capacity-building expectations. The roll-out of the project relied on the successful navigation of complex power-dynamics arising from both the locus of the funding and the research contexts in which the activities occurred. It highlights the need for the challenges relating to capacity-building activities to be more rigorously discussed and supported. Relying solely on individual researchers with no prior experience of capacity-building activities is inappropriate and could undermine both research and development agendas. Indeed, shifting responsibility for mitigating research neocolonialism onto researchers undermines both the efficacy of the research-as-development agenda and brings into question the extent to which decolonization is a priority of funding councils and the governments that fund them.
In response to these challenges, there is growing global recognition of the need for governance and for educating researchers and innovators to responsibly produce technology that benefits society. Issues of participatory research and capacity building are brought to the fore in projects aimed at developing usable technology for LMICs. There is a fine line to navigate between cutting-edge research and "tech dumping" by producing technology unsuited for LMIC contexts. Nor is it appropriate to place the entirety of the responsibility on the HIC researcher to ensure that the resultant technology is sustainable in situ after the close of the project. What is apparent, though, is the urgent need to find a workable balance to avoid the neo-colonialization of technology research that it could unwittingly reproduce or reinforce relations of dependence between HICs and LMICs.
Funders and the institutions that receive funding need to make 59 The OLI project collaborated with UK NGOs "Tech for Trade" and "The Centre for Global Equality" to assist with the development-related aspects of their work. 60 Franzen, Chandler, & Lang, op. cit. note 40. better provision for training to ensure that researchers are equipped to deal with the complexities of these problems.
The recognition that ethical frameworks travel and evolve between research contexts is an important message for both GCRF and RRI discussions. In HIC-LMIC collaborations the most responsible practice may sometimes be less about enforcing practices informed by ethical expectations, and more about being able to relinquish control. However, if research and innovation are understood as collective activities with uncertain and unpredictable consequences, 61 how best control is negotiated requires far more discussion.

ACKNOWLEDGEMENTS
Richard W. Bowman is grateful to Prof. Jeremy Baumberg, the grant Principle Investigator, for his mentorship and oversight. The OLI team would like to thank UK NGOs Tech for Trade and the Centre for Global Equality for their advice and assistance before, during, and after the project. The authors would like to thank Dr Michael Morrison for his comments on the manuscript.

STICLab/Bongo Tech and Research Labs is a private company in
Tanzania and is currently selling OpenFlexure Microscopes.