What are the barriers to equipment sharing? 1. Geographical constraints, as it is clearly more convenient to walk over and use a piece of equipment that is on one’s own site than to travel large distances to use it elsewhere, especially if it is only for a short period of time. Sharing over large distances requires more planning and can waste time when using highly sophisticated equipment which may experience significant periods of ‘down time’. In addition, there will be equipment that needs to be close to other equipment (e.g. to transfer samples), and then it is not always practical to work at a distance. 2. Equipment that is potentially suitable may not actually be available if the task to be performed is very different, or uses different materials, as a result of which the equipment needs to be adapted or re-configured. There are issues with many pieces of equipment to do with training, sample preparation, tuning the equipment and then resetting to the original user configuration, etc., which mean in practice this can only be done for intensively-managed large facilities, where there is a technical support team, a booking system, and a mechanism for recovering costs. 3. The cost of running equipment (i.e. management, service, development, staff time, and consumables) is borne out of the home university laboratory, and external users may not be able to pay the full operating costs. Most universities will be happy to let someone in where they can get something out of it – either running costs for the time used – or a meaningful scientific collaboration with very good prospects for worthwhile and measurable outcomes (e.g. publications in well cited journals). It is difficult to justify doing so if it will cost time, running costs and effort with no compensation. This is especially the case, when equipment that is already well utilised at the university creates problems of over demand, where waiting times are long. 4. There is also the problem that many subsequent grants require deliverables that depend on the availability of equipment already purchased through previous grants. Failure to meet those deliverables (if compromised by equipment being used by other groups) then penalises successful groups in terms of further grant funding. The worst situation will exist in universities with a few very strong groups which have earned sufficient grant income to have built up an effective infrastructure base, with other departments or universities now able to demand access and so reduce the effectiveness of successful groups by forcing them to help support less able ones. 5. Research groups, aiming to attract funding to purchase or run their equipment, keenly solicit interest in the application stage. However after EPSRC funding is awarded, no clear procedures are set-up, and there is little or no effort to accommodate external users. Research groups simply continue collaborating with their own circle of established collaborators. Hence, any scheme should define clearly in advance how access is to be administered (e.g. defined percentage of time, independent advisory panel judging applications). This is more likely to be effective for larger facilities but is necessary at some level in all.

6. There can be a quality issue whereby a group that considers itself as superior to a potential user may be reluctant to allow that user access to the equipment. 7. If the intention is that academics can bid to access equipment based at other (local) universities, instead of purchasing and installing new equipment in-house, then they will need a detailed inventory of what equipment exists, where it is located, how much it is used, and whether it is able to perform the tasks that they want. Creating such an inventory is likely to be a substantial task, with a considerable cost attached. Even within a single university it is not straightforward. Experience of sharing equipment that exists around a university shows that the time taken to locate/request use of/negotiate terms is both costly and a deterrent to busy academics. Things are even more complex where arrangements have already been made with outside commercial partners. Similarly, there may be contractual barriers to sharing existing equipment purchased with funds from other sponsors. 8. Accidental damage (and contamination): there is bound to be a risk of damage or loss if equipment is either transported, or shared with external groups visiting a facility. Unlike a well-supported central laboratory, university groups may not have the technical knowledge (or indeed finance) to cope with damage resulting from sharing equipment. 9. Perhaps the most difficult barrier is the provision of appropriately-skilled people to assist with experiments; if group A needs access to machine X owned by group B (because the research funding is unable to provide A with its own machine), then its unlikely that group A will have appropriately experienced technicians in place; hence on arriving on-site at machine X, it is not clear how group A can exploit the equipment to its fullest without the services of group B's expertise: group B may be unwilling to provide this for financial reasons, or because the staff cannot be spared, or even because it might lead to assisting a competitor group in a sensitive research area. 10. Some equipment sharing in which there is clearly no risk of IP or know-how leakage is hampered by the need for formal NDA/material transfer agreements which are more time consuming then the actual experiment. 11. Another major issue is of timing. If one group wants a particular piece of equipment, it is unlikely that others would necessarily want the same equipment at the same time. Therefore, they would need either to introduce a rule insisting on collaboration before an application or the application of retrospective sharing agreements, which are difficult to deliver because the equipment host already has a strong sense of ownership.

How do these differ for different types of equipment? 12. It is probably most difficult for medium-sized pieces of kit. Relatively small equipment can be shared often on a quid pro quo arrangement. Large equipment (or facilities) will have formally identified costs including those of trained staff to operate or oversee usage, maintenance, costs of hosting, as well as a mechanism for gaining access. For medium-sized pieces of equipment, it may not be appropriate to have dedicated technical support; hence much of the costs in time and maintenance could fall to PIs, PDRAs or students which would be far from ideal. There will also be a difference between equipment which is essentially used in the same way irrespective of the

'sample' or 'problem', and equipment which needs to be set-up specifically for each experiment. 13. Sharing makes most sense for the most expensive highly specialised pieces of equipment where only infrequent access is required and individual use cannot be justified in a local laboratory, e.g. ultra-high resolution TEM and electron beam lithography, and possibly focused ion beam technology. Such equipment also has significant running costs, and there is a need to finance replacement parts on a regular basis. Other pieces of equipment are better placed locally (e.g. routine -electron microscopy, and high-performance computer equipment), especially if they are sufficiently utilised to warrant their existence in the local laboratory; independent arbitration may be required to decide usage. 14. Portability of equipment: large, fixed items (electron microscopes, etc.) need to be used in situ, but smaller items could be crated up and transported, although that would reduce the usage efficiency as well as introducing a significantly larger probability of breakdown. Future grants should have such costs incorporated within the funding envelope (along with the requirement to share), but existing facilities will need a central fund to offset transport charges.

How could the barriers be overcome and by whom? 15. By the provision of equipment access schemes which have previously been used by EPSRC to encourage universities to open up top quality research facilities to others, by paying the running costs associated with providing access to external users. Facilities will only open their doors if there is a financial incentive. EPSRC should put real money into such schemes, and demand that clear routes and procedures for access are defined. Given that there are several top quality facilities in the UK, EPSRC should initially focus on opening up access rather than offering schemes to purchase new equipment (in the past often duplicating what is available elsewhere in the UK). 16. Having a scheme which doesn’t require a lengthy and detailed application process for funding approval, i.e. something that is agile. There is a need to run an efficient booking system. This would need to be properly costed, and is not obvious how this would fit into the fixed term grant funding model. 17. Introduce incentives for sharing at the grant application stage. Where a major piece of equipment is requested, there could be a requirement to try to identify potential sharers in advance. 18. Universities need to learn to act more in a regionally or nationally co-operative way. They must also develop charging models that treat both internal and external users equally and fairly. In most cases, developing infrastructure at a regional level would make most sense as this minimises the distances that would have to be travelled. This will require institutions to set aside the ‘university reputation’ agenda in terms of a more ‘regional excellence’ agenda. Universities will also have to work closely together to obtain funding for state-of-the-art equipment with the understanding that the choice of location for it could be a decision to be made by funding bodies. A positive spin-off of this could be a strengthening of collaborative research between UK universities; a

possible model is where a group of universities find common cause in hosting a range of different facilities, taking it in turns to ‘benefit’ from winning each. 19. A very comprehensive and accurate (web-based) list of equipment would be very helpful. It would need to be set-up and maintained by the research councils with the full co-operation of the universities, and cover at least equipment purchased within the last five years. Perhaps some of this equipment may not be needed now by the original PI, and so the sharing may be welcomed. However, it may be difficult to force academics to subscribe to it, especially in the case of non-research council funded equipment; an additional thought on this, is that if a grant application requests new equipment, will the proposer(s) need to show that existing equipment somewhere else should not be used?. What might work is to allow apparently unrelated grant applications to bid for new shared equipment, so that the sharing arrangements can be in place before the purchase. There is a reasonable analogy with the use of the Grid to share computing resource. So in principle high-performance computing might be a good place to try out any new scheme. 20. Existing owners of target equipment must be reassured that sharing will not be a cost on them, either financially or as a research penalty. Moreover, potential users must also be assured of adequate technical back-up on site, and timely access to facilities that makes sense for their research programme timelines. For equipment sharing targeted at relatively small, portable pieces, it might be possible to co-ordinate experienced technical help via the central facilities. For equipment that must remain in situ at an institutional site, some compensation for the hosts has to be organised. 21. Some practical difficulties could be alleviated by providing a ‘service’ facility where users post/ship samples and the work is done at the facility and the results sent back. This of course is only possible in a few areas. 22. Increased use will help to bring the per diem costs down, but abandonment of the principle of charging for depreciation of equipment and buildings would be a useful step. RCUK should be encouraged to consider applications for equipment purchases on the basis of a scientific case, rather than there being a built-in assumption that researchers will replace like with like no matter how the research has changed, when old equipment fails, which the depreciation model will tend to favour. 23. Standard IP agreements which protect know-how without overlong negotiations would help. Collaboration rather than simple equipment-sharing will be preferable where novel equipment and processes are central to a research programme. More generally, RCUK could generate standard equipment sharing agreements covering the issues identified.

The Institute of Physics is a leading scientific society promoting physics and bringing physicists together for the benefit of all. It has a worldwide membership of around 40,000 comprising physicists from all sectors, as well as those with an interest in physics. It works to advance physics research, application and education; and engages with policy makers and the public to develop awareness and understanding of physics. Its publishing company, IOP Publishing, is a world leader in professional scientific communications

76 Portland Place London W1B 1NT Tel: +44 (0) 20 7470 4800 Fax: +44 (0) 20 7470 4848 Email: [email protected] Website: www.iop.org Registered Charity No. 293851