I am interested in submitting a proposal to the NSF digital government initiative in the area of crisis management with a particular focus on research relevant to systems that would support the collaboration between civilian and military authorities. We have already been involved in this area, providing support to the XII (eXtreme Information Infrastructure) initiative led by Lois Clark McCoy from NIUSR (National Institute for Urban Search and Rescue http://www.niusr.org ). Her impressive team already includes important players including DISA, DOMS-MSCA (Director of Military Support -- Military Support for Civilian Agencies), FBI (responsible by Presidential decree PDD-39 for terrorist induced crises, with FEMA following in consequence management). Our last demonstration (in May at Hanscom Air Force Base) included operational emergency management units from BEMA and MEMA (Boston/Massachusetts emergency services). I think that this is a good seed for a testbed that spans first responders (firefighters in the field) to command centers, domain experts and a rich computational environment. Other agencies such as the Public Health Service and Center for Disease Control would be involved in the terrorist biological events addressed by XII. We suggest Los Angeles as an attractive testbed for we already have ties at a number of levels there. Lois and I participated in the workshops that led to the NRC report on Crisis Management and there have been several relevant studies of the importance of our chosen focus, which is a major responsibility of the future military. This is for instance articulated in the Report of the National Defense Panel. "Transforming Defense -- National Security in the 21st Century" December 1997. (http://www.dtic.mil/ndp/). I expect to get substantial industry involvement from companies that span those that build military command and control systems to the smaller ones supplying the civilian crisis management software and systems markets. I would also expect the insurance industry to be very interested. Other relevant government involvement will include the National Guard, the Coast Guard, Marine Corps and the Department of Energy. In the latter case, I would involve Los Alamos's new Delphi effort on time-critical predictive simulation under Andy White who speaks to this point at the Grids 98 workshop July 27-78. Our NSF proposal would focus on the research issues underlying this problem area and I see that this could be a wonderful driving application for the high performance computational/information grids, which are now being developed by NSF DoE and NASA. This challenges collaborative, database and computational technologies and your call for proposals in fact lists some very relevant areas as did the original NRC workshops and the recent Schorr-Stolfo report underlying the NSF solicitation. Interoperability between expensive high functionality military systems and the more ubiquitous low cost civilian systems is an essential requirement where standards such as XML and CORBA can be investigated. So I first want to check that such a proposal could fit your guidelines. Further I would suggest a two-phase approach. Although Lois has put together a good project team at both technology and user levels, we currently do not have a research component for the XII effort. So phase I for FY98 funds, would be an effort to set up the research activity and develop the architectural framework coming out of the set of further planned demonstrations. In particular we will identify some specific projects to provide focus for the initial research. These would be one or more components of the Crisis Management process -- a typical suggestion is the study of the mechanisms that allow the Pentagon access to local and state geographical information systems. I will work with the NCSA Alliance, Los Alamos and other players in the grid area to understand how to set up a crisis management testbed that is a major driver of and benefactor from the new high performance grids. Note that one controversial issue is the use of a distributed grid for time-critical simulations; we would research this but note that one can start with a collaborative information grid in which individual components are parallel simulation engines. These phase I studies could lead in phase II to one of the center proposals you identified for FY99. I thank you for any comments on these very preliminary ideas.