Saturday, April 3, 2010

Notes from yesterday's crisis-mapping meeting

The good stuff is all on the Facebook group now, and soon we'll have a dedicated Jor El website. I'm just posting this here so that we have a non FB backup for now.

Here's the plan from the crisis-mapping meeting yesterday (April 2) with Aysegul Duygu Saylam (of Istanbul Twestival), Asli Suner, and Bayram Yildizoglu. Knut Milbrendt phoned in from Germany.

1) No point waiting for help setting up the platform. I'm going to see if it's really that simple to download and customize the USHAHIDI platform, using my PC as a local host.

ADS has a friend who can help me with this if it all goes pear-shaped. I seem to have committed myself to doing this over the first genuinely glorious spring weekend of the year. It seemed like a good idea at the time.

If all goes well, this will be up and running by Monday. Then we can start playing with it.

Obviously, this isn't a long-term solution, not least because all systems we'd need in an earthquake need multiple backups, but it will give us a demo and allow us to experiment with the platform.

We don't expect to be using Twitter as the primary source of input data: That will come from SMS messages. But until we've made arrangements with the cellphone companies, we'll be testing it with Twitter, and certainly Twitter will be an extremely important source of information, if Haiti and Chile are anything to go by.

We're then going to test it in three phases:

1) We're going to do a Twitter test run next week, or as soon as the platform is reasonably bug-free. ADS will publicize it on Twitter. I'm asking Ali Orcun Guner to get us as many student volunteers as he can from Bahcesehir University to geolocate and input the mock-Tweets. Obviously, the Tweets need a very clear code to indicate it's a test, not a real emergency.

2) If this goes well, we're going to show it to AKUT, which has expressed great interest in this idea. Then we'll schedule a second-phase test, again using Twitter, in which AKUT will use the system and send real rescuers to the sites. They'll let us know if the system seems valuable to them and what could be done to improve it.

3) We'll then have something concrete to take to cellphone companies and people who can help us to publicize the system. The goal is to make this system extremely well-known PRIOR to an earthquake, and to have a large cadre of volunteers trained in using it, so that it can go into action immediately and smoothly the moment it's needed.

To that end -- again -- it's essential that there be a well-publicized SMS shortcode that everyone in Turkey knows to use in an emergency. It does a lot less good to publicize this number afterwards than before. People also need to know how important it is, when using this number, to be very specific about where they are and what they need. This was one of the key lessons of Mission 4636 in Haiti. Far too many heartbreaking SMS messages didn't include the information rescuers needed to help the people who sent them.

The platform will be used to two ways:

First, to direct emergency local and international rescuers after a quake, in what we know will be a highly chaotic situation.

Second, it will be used to map hazards prior to the quake and give people practical information about how to prepare. The SMS shortcode will allow people to phone in issues of concern: (eg, "I'm worried about the school my kids are in, what should I do?") and will direct them, via an automated message system, to practical advice and resources. This will give us a better picture of which areas are really problematic -- one independent of government estimates.

Using this data, we can predict where emergency assistance will be needed (and obviously shorten the time required to get it there.)

It will also be a way to let people know about the practical steps they can take prior to an earthquake to increase their chances of surviving it. This information is right now not well-understood here at all.

A few more points came up:

BY is concerned that Jor El needs some kind of official status, as a destek or NGO. He's worried that the people whose support we most need will ask, "And who the hell are you?" Likewise, there are a lot of legal requirements we'll need to meet if we're to do fundraising.

AS mentioned that we could possibly simply attach ourselves to MAY, but BY pointed out that MAY's connections to the CHP might result in the appearance that Jor El is partisan. This is a real concern, and gave us all pause.

I'm extremely leery of the hassle and paperwork involved in becoming some kind of official destek or NGO -- I like to think, in my boundlessly optimistic American way, that the answer to the question "Who the hell are you" should be, "We're a bunch of people who met on Facebook, but look at these superb ideas we have and what we've done. Aren't you thrilled? Now work with us." But I do realize that sometimes things aren't that straightforward in Turkey.

We could use some good advice about this problem. We need to get Duygu Erten to join the conversation -- this is her field of expertise. Ayse Sumer, maybe you have suggestions as well.

Finally -- NB Hilary Chan -- Knut mentioned that Jen Friel, who runs the blog "Talk Nerdy to me, Lover," has devised as system for mirroring and backing up posts and comments to Facebook. If we could back up our Facebook site this way, it would obviate one of our big concerns -- that all of this will disappear in a Facebook glitch. HC, perhaps you could investigate?


  1. Crisis Mapping can be described as combining the following 3 components: information collection, visualization and analysis.

  2. Learning to fight is a good way to fight for yourself.

  3. Crisis Mapping can be described as combining the following 3 components: information collection, visualization and analysis.