[OpenRelief Outreach] [wargaming] Situation simulation to allow understanding of OpenRelief and optimal decisions for deployment of OpenRelief solutions (was: Re: Wargaming for Disaster Management Simulation)

[Shane Coughlan]

Hi Kevin

Thanks for offering to assist in developing a situation simulation to help train people quickly.

It would be good if we could reduce the simulation to just the pen, paper and guidebook. Our problem and our proposed solution is actually pretty focused, and will not cover general disaster relief.  Let me explain.

The OpenRelief tools won't work in isolation; the modules and UAVs are just collecting and sending information back into disaster management systems like Sahana Eden (see: http://sahanafoundation.org/products/eden/) that deal with the actual response strategy. Dealing with infrastructure restoration (roads, power, water) is meanwhile taken care of via local and national guides in each country (see, for example, this overview of local government earthquake disaster management in Japan: http://www.hyogo.uncrd.or.jp/hesi/pdf/peru/saito.pdf or this paper covering 'Recent Trends in Earthquake Disaster Management in Japan': http://www.nistep.go.jp/achiev/ftx/eng/stfc/stt019e/qr19pdf/STTqr1907.pdf). 

For this reason, it would be best to keep any simulation focused on the projected launch modules for OpenRelief (UAV, modules for seismic, radiation, submersion, weather and FM transmission) and how people can deploy them usefully. Our goal is to help people use OpenRelief tools to turn uncertainty into certainty as they implement their own varied plans for recovery and restoration of infrastructure. That's pretty important in itself, but it does limit the scope of our problem and our necessary simulation.

Regarding the metrics for success, the situation is almost entirely unknown during the first few days of a disaster. OpenRelief tools can be deployed to obtain awareness around areas where (a) people are/may be located like shelters and remaining houses, (b) areas which will be used in logistics like remaining elevated areas around towns, and (c) dangerous items like failing reactors are located.  The maximum information obtained in a given situation with the available modules is our key metric. Assume zero (or near zero) data and try to obtain the best fidelity with the available OpenRelief tools (plus any other tools to hand, but we can't count them in our simulation). 

That could work in two ways, with one being a general overview simulation that takes very little time ("what are these UAVs and modules for, anyway?"), and another being more detailed situation simulations to help people answer questions like "how do we get information and initially monitor conditions in ten villages across a 20km stretch of coastline?" The first use case could assume that quite a few OpenRelief modules are available. It's a "best case" simulation to contain a general disaster. The more detailed simulations could assume that in practice the available UAVs and the available modules will be limited. Therefore the required area versus the possible module coverage could inherently lead to imperfect information. The decisions therefore would seek to obtain the maximum possible information given these constraints. 

When it comes to simulating these activities, it would be nice to see if game theory could be applied (see my previous email) as that would fit into existing wargaming practice and make adoption easier. It might also be useful to explore if the decision process could be based around OODA (observe, orient, decide, act), as this will fit right into existing civil and military decision-making processes. 

Regards

Shane

OODA loop:

Source article: http://en.wikipedia.org/wiki/OODA_loop
Image location: http://en.wikipedia.org/wiki/File:OODA.Boyd.svg

On Mar 21, 2012, at 2:49 AM, Kevin Walsh wrote:
> Shane has contacted me with regards developing a Wargame to help demonstrate the tech, you guys are providing, and to help people who are in charge of such decisions, understand better the effects of various methods of deployment and use.
> 
> It's a big challenge, there are so many variables, and Its hard to say what exactly your modules will do or how they will perform under adverse conditions, but I think I have a grasp on the key feature that will become the core of the game mechanic: Information, or rather the lack of it. The big issue during a disaster is the difficulty of forming a comprehensive picture of what is going on, and therefore make good decisions about what to do about it. The tech your developing seems to be aimed at improving the information available to make these decisions, so the key feature of a game meant to show case their use will be limiting the information that the player has about whats really happening, and depending on what assets were deployed and how they were utilised, the player gets more information back and can therefore choose to react to the crisis.
> 
> Ideally I would like to produce a text based game computer game that runs a detailed and multi layered simulation behind the scenes and then feeds information to the Player via a Sahana Eden Interface so the player would be in the position of a real crisis management professional.
> 
> How ever in the short term I'm working on a pen and paper role play strategy game, for 2 players, utilising a real world map, pen's, pins or counters, dice and a stack of debating cards.
> Player 1 would act as the Disaster Relief Manager, and Player 2 would Act as the Environment. The two Players would divide up the map into districts and write down the features of each district using a guide int the game book and their own knowledge of the area, then Player 2 would choose or randomly determine the type of disaster while Player 1 was deploying the assets available. Once ready Player 2 would add the effects of the disaster to the back of each districts cards and then Player 2 would attempt to assess the situation with what ever assets were in play Player 2 feeding information as it was discovered. as time ticks by Player 2 would continue to modify the situation to take into account ongoing problems caused by the initial disaster while player 1 would attempt to provide Aid to the victims and Re-establish infrastructure such as power, phone lines, water and roads. 
> 
> A key feature that I have yet to define is Metrics of success, so that one run can be measured against another and the difference between the Strategies can be clearly seen.
> 
> From you guys I would like links to information on the details of disasters and their effects, the survival rates for post disaster refugees under different conditions etc, and also an assessments of the types of assets that Aid Agencies already have as well as assets that you guys are making or even just thinking of making. 
> 
> The more I know the more I can Include.
> 
> Kevin Walsh
> _______________________________________________
> Outreach mailing list
> Outreach at openrelief.org
> http://openrelief.org/mailman/listinfo/outreach_openrelief.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://openrelief.org/pipermail/outreach_openrelief.org/attachments/20120324/bcfaa659/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 800px-OODA.Boyd.svg.png
Type: image/png
Size: 67408 bytes
Desc: not available
URL: <http://openrelief.org/pipermail/outreach_openrelief.org/attachments/20120324/bcfaa659/attachment-0001.png>