Notes on Tech Transfer Mechanism: These notes are concerned with a Vision or modus operandi for the whole tech transfer process. The main pont is about putting people-in-place with technological gizmos like video conferencing and the world wide web as valuable support mechanisms but not replacements for people. My main concern is for *how* the people in place are put together. Although NPAC is probably more flexible in regard to its staff than other sites, it is still hard to envisage a mechanism for staff to be uprooted and physically dispatched around the country for substatial periods of time. Some new (junior) staff can be recruited with this requirement wired into their job description, but you still want your experienced staff to make the rounds too. Some sites will only have post docs available, others will have senior staff. How do we customise things so all the SRC sites feel equally loved? How do we ensure that we are not seen to be fobbing off the MSRC sites with junior unqualified staff? This problem is well known to me as I had the exact same problem at Edinburgh Parallel Computing Centre. Our customers wanted staff in place and although this was cheap for us (as it would have lowered office and facilities overheads costs) very few staff wanted to be located at some Military installation or at some company's plant for an entire year. We used to get around this by using the rotating team mechanism outlined below. We would arrange for a staff member to stay for 2-3 months at a time, and rotate 3-4 staff this way. This actually works better, since all staff still feel atached to their home organization, are abl;e to kep up to date with events/technologies there, and ae able to operate better in the field calling on expertise they don't have directly. I suspect that the MSRC sites will be happy to poach some of these staff in place in the long term. (That certainly happened to us). We have to acknowledge that this will happen and indeed is one of the long term tech transfer processes. What I believe is fundamental, is that we think through the manpower issues now, and that we make some attempt to plan possoble career routes for these staff in place. Some will indeed wish to move to MSRC sites, but others will need sufficient links with their home organisations so they still feel employed there at the end of their posting. One way to explain the distributed team approach to the MSRC sites would be to illustrate in terms of a CRAY or Thinking machines site engineer. These companies put this person in-situ as part of the package of buying one of their machines, although it would not necessarily be the same person in perpetuity. The point was that the site engineer was familiar with his suppot organisation and could call on all sorts of hardware/software expertise. We should be able to do the same type of job a the applications level. (This analogy should be familiar to the sites who have bought large machines in the past). This also lets us form distributed teams from the universities involved and thus get the best cross fertilisation of ideas and technological details. In other words it can benefit ourselves by letting us form consortium/collaboration teams for our own ends in a way that is synergistic with the MSRC needs. We need to describe the various nitty gritty details such as a meeting to get the whole team for each CTA together periodically, video conferencing, setting up web pages for each CTA with FAQ's and search mechanisms. This coordination can be handled by the coordinator for each CTA. This stuff is easy to write about if we all agree on the basic staff-in-place mechanism. The following words might be appropriate for the proposal itself: ------------- Fundamental to this proposal is the technology transfer mechanism using personnel in place. This can be accomplished by employing staff who can be physically located at the MSRC sites for substantial periods of time, but who can draw on their own home organizations for support and expertise. These staff-in-place will also be familiar with the appropriate circle of ``preferred vendors'' of expertise. We believe that the technology transfer needs within an area are in fact beyond the abilities of an single individual, therefore we propose a distributed support team that can be customised for particular needs of individual CTA's and MSRC sites. We envisage a team of 2-4 individuals for each CTA, with a central support team member who will be physically based at the appropriate lead CTA (university) site and who will coordinate activity with the backup and other university sites and teams memers as well as staff in place at the MSRC sites. Some flexibility will be needed in assigning the other 2-4 team mambers. These staff will be drawn from staff employed by the lead, backup and other university organisations and will be physically located at MSRC sites for perhaps 6 months a year or more on a rotational basis. It should be emphasised that this teaming arrangement allows coverage of CTA's that apply to more than one MSRC site. It also allows some flexibility in the grade of personnel that can be assigned. Skilled personnel in each of the CTA fields may be scarce and not geographically mobile. Our proposed support teaming mechanism allows us to dynamically reconfigure the staff in place and to have a senior staff member on-site for some short period of time, with a more junior member there for the remaining time. [Joe - which CTA's map to which sites? I guess this is your concern as prime contractor?] A considerable range of technological support mechanisms can be brought to bear to assist the teams in place. We are familiar with video conferencing as well as the use of hypertext and the World Wide Web as ways of enabling physically disparate team members of working successfully together. We use this technology ourselves in technology transfer activities between industrial and academic members of our InfoMall technology transfer program and we have also successfully demonstrated this technology to the US Air Force Material command as a mechanism for building decision support and technology transfer systems. In spite of the utility of this technology we recognize the value of staff-in-place and envisage our proposed distributed team as a mechanism to make it work in practice. Our experiences with in-depth technology transfer to industry and the military suggests that the diverse requirements can only be filled by a customised distributed team with the technological support facilities of the World Wide Web. [ We should describe emerging tools such as Hyperworld and any NCSA developments here, and statet that the technology transfer process links with tool development elsewhere in the proposal ] --------------- Now down to details, and costs in particular. Not all the CTA's require the same degree of support, and indeed some overlap in expertise. For example, I believe the knowledge necessary for Computational Fluid Dynamics overlaps partially with Computational Electromagnetics and substantially with ground water simulations in environmental modelling. Taking the overlaps aside however, suppose each CTA need 2-4 staff member as described above. I would expect that the lead site might have one senior staff member in coordination mode at its own site and one or perhaps even two staff on-location. The backup site might have one or perhaps one half of a full time staff member available for location work. ``Other'' sites might provide resouces on a per demand basis or might perhaps have half a full time staff allocated. Exact costs would depend on the staff seniority involved, and just what mechanism we need in place to cost a staff member in-situ. What level of compensation is due to a staff member on-site? Does this get factored into the travel budget? As an example, NPAC staff cost approximately $65k per annum for a junior (ie MSc level) staff member to in excess of $100k per annum for a senior (Phd++ level) staff member. On the basis of the scenario above: a CTA support team would cost: 1 Senior Staff coordinator 100 1 Senior Staff traveller/on-location 100 2 Junior Staff on-location 130 Travel (based on 50% of travelling staff) 115 Giving a total cost per CTA of $445 per annum. This is the scenario I would imagine for a CTA like FMS which has very little technical overlap with the other CTA's. Perhaps the other CTA's could be reduced somewhat. Also, this takes no account of special facilities required. For example, video conferencing costs and a dedicated World Wide Web server.