For many years have researchers and R&D-people in industry done impressive work to further advance our understanding in how to design innovative and effective IT-support for emergency and disaster response work. Some of these solutions have actually been evaluated before they have been transformed into products. The evaluation has often been done by having domain experts examining the solutions and based on some heuristics judging the feasibility of the specific solution. In other cases have the evaluation consisted of a series of fieldtrials and extended test-use in real or as close as real situations. Sometimes have the proposed solutions only been evaluated against a made up worst-case scenario that has little or no link to real world situations.
Sadly, none of the above strategies give us any proper insights into what really makes a difference. The extended test use is probably the best strategy but it also means that we have a significant cost in terms of training and integration which most often make this approach to costly or complex to accomplish.

I believe that it is time for as to balance our strong focus on the situated use and various work practices and complement with an approach that is instead based on models and simulations of crisis events. The sad news is that we do not have a crisis dynamics theory which could serve as the foundation for any model of crisis events. The good news is that we have good examples of research teams in the US, Asia and Europe that have done some impressive work on multi-agent simulations that very well could serve as a starting point for further work that in the end could result in models that can capture the dynamics of crisis events.

What I believe is necessary, is to start exploring how we can model a crisis and the dynamics of a crisis without ending up in specific organizational aspects. My idea is that if we treat a crisis as a natural phenomena and in an actor-network inspired way treat human-actors and non-human actors as equals, then a model could perhaps be constructed.

Would it be possible to identify the general actants of a crisis and would it be possible to identify the dynamics of the relations between those actants?
What I do know is that researchers traditionally deeply positioned in the qualitative tradition in social sciences could perhaps benefit of reaching out to researchers that typically use quantitative approaches and working on natural sciences. Perhaps, by organizing a project that provide a small but fruitful area in the intersection of these two worlds, a model could be materialized.

I might be a bit off the grid with my thinking here. If you think so, please let me know about your approach to explore this topic.