I’ve corresponded with Mikel Maron for a while now, interested particularly by the work that he’s been involved with introducing Wikis into the UN. However his first love is geospatial and his favourite project is OpenStreetMap, which is a free editable map of the whole world that can be viewed, edited and used in a collaborative way from by anybody, anywhere.

Mikel has been thinking about how this type of approach might be used in disaster response, and he recently gave a presentation at the State of the Map event in Manchester in which he outlined his thoughts so far - you can hear a podcast of his talk, view the slides that accompanied it and read the notes from it. Mikel acknowledges that he’s not an expert, and while there’s a lot of things that I agree with him about, there’s also a lot of assumptions built into his talk.

From Mikel’s notes:

In 2005, the Southeast Asian Tsunami caused destruction on a scale beyond comprehension. And Katrina later in the year showed that no country, however developed, is beyond disaster… And my thought, and many people’s thought, was that these shiny tools from the web were ideally suited to a distributed response to disaster, where a hierarchical one was failing.”

The assumption that the usual (”hierarchical”) response to these disasters was failing is quite a big one. The Tsunami Evaluation Coalition found that the initial response to the tsunami by the international community was actually quite effective - it was the longer-term reconstruction work that has failed. I will agree that the hierarchical federal system in the US clearly failed during Katrina - and although its defenders will argue that the failure was political, and that the system itself works just fine, I would counter that any system that doesn’t take into account the politics of disaster response is simply setting itself up to fail when a large-scale disaster strikes.

One thing that is true, however, is that in any disaster, most of the support is provided at the local level by local actors, often outside the formal response itself. We’ll get back to that, but the main gripe that Mikel has - and it’s an entirely valid one - is that data is frequently inaccurate and out-of-date. It’s worth quoting this part at some length:

My friend Jesse Robbins… headed down and helped lead the set up of a relief operation, not too far from where this bridge on US Route 90 had been completely destroyed. However, the Red Cross was giving evacuation directions to cross this bridge, so loads of cars would stop at the edge of this pennisula with confused drivers. Jesse phoned the Red Cross multiple times to complain the bridge wasn’t there anymore .. and they responded “but that can’t be — it’s still in Google Maps!”

The root problem was that the Red Cross was giving out inaccurate information based on what they thought was a credible source: Google Maps. It’s not the job of the Red Cross to maintain accurate datasets on road infrastructure, but they do have an obligation to be reasonably sure that the information that they provide is reasonably accurate. Naturally, over-stretched Red Cross volunteers were going to rely on whatever they could find - in this case, Google Maps, which is a perfectly reasonable choice.

On a practical level, Google Maps will be unable to update their data as quickly as disaster response requires, simply because that’s not the way in which maps gets updated. No matter what source Google Maps uses, all of these magical online map services rely on the not-so-magical process of somebody actually doing the mapping, and to confirm a major change such as the destruction of a bridge requires an official survey. (Google Earth is a different matter, as remote sensing imagery can be updated almost immediately - for the right price.)

Despite the fact that OpenStreetMap can be updated much more quickly than Google Maps - a volunteer could submit a more recent update on the condition of (for example) that bridge - why should I trust OpenStreetMap more than Google Maps? While I might be able to tell if their information is more timely than Google Maps (by checking the date stamp to see when the information was submitted), I have no guarantee that their information is more accurate or reliable than Google Maps, particularly if that information could have been submitted by anybody.

This is because my interest in that bridge during a disaster response is not knowing where the bridge is (the geospatial component) but what its status is. The example Mikel gives is of the complete destruction of the bridge, which is a simple binary choice (destroyed/not destroyed); but what if the bridge is intact but can no longer bear its full load due to structural damage? That might mean it can be used as an evacuation route for cars, but not for heavy trucks bringing supplies into the area. Getting that wrong will just compound the problem, but it requires somebody who knows what they’re looking for to submit the information.

So what does OpenStreetMap need in order to be more credible than Google Maps? It needs input from professionals who can make informed judgements about issues about humanitarian needs, logistical issues, security information - all the things that the aid agencies already do. I agree with a lot of what Mikel says about how an approach like OSM’s can offer more transparency in the delivery aid, which is something we desperately need. By the same token, the hopes of Mikel and others is that technology can flatten the hierarchical approach and help beneficiaries to meet their own needs better (effectively removing the need for outside assistance).

However an open and distributed system is vulnerable to abuse from unexpected quarters. I believe that beneficiaries need to be more involved in gathering data and have more access to data that’s relevant to them - but I’m also aware that they are not disinterested parties. We might think that information submitted by a community leader about the needs of a village affected by an earthquake is going to be more accurate than that submitted by an NGO volunteer - but the community leader has an interest in exaggerating his village needs to maximise the amount of assistance they get, a phenomenon which those working in the field are familiar with.

Mikel’s arguments for OSM (and other distributed approaches) do have a lot of merit, but I think they conflate two separate but related problems. The first is making aid agencies more effective and transparent, and the second is empowering beneficiaries through technology. Both of these are valid objectives that we need to work for, but they might need different approaches to be achieved - while some of my colleagues (such as Michael Howden) believe that Web2.0 (which includes initiatives like OSM) can bridge the two sides, I’m still not convinced.

Of course, this discussion is nowhere near finished yet…