Yesterday, I visited Mukuru slum (still a blank spot on the map) to discuss possible collaborations and methodology sharing with an ongoing participatory mapping project there coordinated by Emory University and Partnership for an HIV Free Generation, at a well apportioned community site built by Micato Safaris.

Mukuru has developed on different lines than Kibera … the road infrastructure is relatively planned, there are designed gutters, the area is flat and sprawling over a huge area south of the Industrial Area. It also receives less attention than Kibera, so there are relatively less NGO and CBO programs trying to fill the service gap. The Mukuru group has used many interesting participatory techniques to engage the community, some based in the work of Robert Chambers, who we met last week at Erica’s IDS talk on Map Kibera (and told us a brief and awesome story of his mugging in 1960s Kibera while rock climbing). Particularly, the group there had undertaken a “community asset mapping” in two villages of Mukuru, a paper based cataloging of community services, with actual GPS mapping done by a consulting company, resulting in a few map prints and spreadsheets detailing services. The resulting data is good, but they complained that it’s not really reusable or updatable by the team directly. So they were very interested to hear about our techniques which both built capacity to create and make use of map data, and other kinds of information, directly … they can do it themselves, have lots of ideas for applications in the community, and perhaps soon we’ll have an opportunity to explore some merging of methods.

On Tuesday, I met with Umande Trust, who do amazing and innovative work in sanitation, best known for introducing biogas latrines to Kibera. I talked with Aidah mostly about data … we have data from our recent water and sanitation mapping, Umande has amazing data in Gatwekera and adjacent villages, exact locations of many standpipes, toilets; and water piping installed by Nairobi Water.

But also we talked about their programs generally. In an innovative compromise to the usual policy of simply ripping out illegal pipe connections in slums, Umande has worked with Nairobi Water for installation of master meters for the entire area, and may work to dialogue directly with the community about how to legitimize and make secure the entire system, in a way that works for everyone. They are also discussing the installation of proper sewerage, partially to help with the volume of waste in the biocentres. Turns out too many people are using the biogas latrines for the decomposition process to fully digest the waste, an example of the challenges of when innovation meets reality.

Umande has also done community asset mapping of the entirety of Mukuru, by actually training community member on GPS … those individuals and that data could be a great start to further work in Mukuru. Of course we need the data shared. I was delightfully surprised and also shocked that I hadn’t seen these maps before. But realize, this is not due to any intent of Umande, Aidah is very willing to share the data for reuse if only there were an infrastructure to use. I’m interested to build off the Digital Gazette, an integration of Crabgrass and GeoCommons for use in Northwest Pakistan, to help build up the information commons of the slums.

So, on Monday, I attended a Nike Foundation event to discuss innovative use of technology to assist young girls. This is something Map Kibera has focused on through our association with Unicef, and the Nike team were very interested to learn and listen, kudos. One point during the presentation touched on Pamoja Mtaani a video game developed by Time Warner that conveys messages about HIV/AIDS prevention. As chance would have it, I got to see Pamoja Mtaani in action directly in Mukuru.

Before our meeting started yesterday, I was given a tour of one of two centers built in Mukuru to “host the game”. My jaw dropped as we entered a spacious room, with 20 gleaming computers showing Pamoja Mtaani. The game looks pretty well designed and fun, definitely innovative and not overly didactic … the game starts with a matatu robbery, includes graffiti and MC games, runs through a virtual Nairobi … a kinda slum version of Oregon Trail.

What struck me cold was that most of the computers were unused, and were completely locked down to only run this game. Here was a resource that any school anywhere in the world could use well, and Time Warner only permits their game, a game that you also can not download and run freely. Frankly, an idiotic implementation of a good initiative, and somehow emblematic of many things wrong with the inappropriateness technology in development. I can understand the desire to make sure the game is actually played, rather than just giving away computers for kids to check Facebook (of course that would happen) … but you do that through a well designed program and continual involvement rather than locking away all the other potentials of this computing resource.

Now we could make pretty good use of those computers for mapping in Mukuru. As it is they are useless for anything else.

At the other end of the spectrum, a few days before, I was reminded of the Hole in the Wall computer project, when Tanya Notley asked me my impression of this TED Talk on education innovation in the slums. The Hole in the Wall computer, simply an open kiosk installed in the middle of the slum, open for children to experiment on without restriction, is one of my favorites. The radical experimentation here has been widely celebrated and hyped, including in that TED Talk.

Now someone recently went back to actually visit these Hole in the Wall computers. The program had change considerably. Wonderfully, there are many more computers in the slums … but they are not the unrestricted free for all the original program experimented with. Apparently those original computers quickly stopped functioning, and the program now looks like a more traditional digital village, with computers, internet, and programs to use them. There’s still space for exploration, but not without supervision.

So in response to Charles Leadbeater, I wrote Tanya..

I broadly believe in what he’s saying, but the talk is pretty hand wavy. The critique of our education system is not new, and plenty of examples of experiments from the past 50 years abound. Certainly there is innovation in education happening in marginalized areas, enabled by new technology, access, and approach … we see ourselves fitting this movement quite well. What’s more difficult is to see how informal education will connect with the existing models … which are still essential for legitimacy and maximum opportunity for all. We’d like to see the opportunity for mappers to continue studies in GIS if they are keen, and several are, but there is no consistent financial aid here. Seems to me that there’s a need to still engage with traditional institutions, open up the way they approach education.

So maybe after some more research, he’ll have some insights! For one, he needs to ground check his facts … just yesterday I read that the hole-in-the-wall computers more or less stopped working 3 months in, and they’ve now adopted a more traditional approach. I still love the audacity and thought provoking of that project, but too often in innovation (and generally) we’re hiding the failures and not learning.

Perhaps one of the hardest lessons I’ve been learning is the limits of innovation. Map Kibera is innovative, and it’s much hyped, but is it in danger of being another project that only looks good from afar? To really make an impact with technology, is requires far more than simply doing something new. It takes a lot of work which you might think is boring, lots of discussion, lots of program design, lots of failure and revision, lots of reality. There are limits beyond simply getting computers and internet into marginal places, limits beyond training, that have to do with the dense interconnection of all issues facing our increasingly urban and marginal world. The shiny glean of technology starts things off, but after that the work may be the same as ever … learning from each other, respecting different points of view, long negotiations of how things can change for the better.

Response from the international community was that any mapper has the choice to map whatever they like, especially if it physically exists. Deleting someone else’s data would be against core principals of OSM, of open data.

As Frederik states, this puts us in an ideological paradox

With Russia we now seem to have a case where it is possible that a
majority of the local community decides to work differently from the
principles that we have grown to cherish, namely letting everyone map
what the hell they want. We now have people who not only want to keep
certain things out of the map, but even delete these objects where they
have been mapped by others. It is even possible that they form an OSMF
chapter supporting these actions.

If this happens, then we have a clash of two principles. Either we say
that the principle of subsidiarity overrides the “freedom to tag”,
allowing the Russians to restrict that freedom in their area because it
is, after all, their area; or we say that the “freedom to tag” is the
overriding principle and we will not tolerate any local community to
tell us outsiders how they would like to be mapped.

This could have repercussions for mapping in many restrictive authoritarian regimes. It’s my position that for the vast majority of cases, open data is better and the default, and in fact, makes us more secure and informed. This was one critique of our mapping in Gaza, that showing details of Gaza would make Gazans more vulnerable. Fortunately, Gazans also agree that information is all about changing perception.

In general, I view these edge cases as a question of power. Hiding information protects those already in power, but not those that are already marginalized. Legitimate cases to me is only information that puts dis-empowered people at risk, such as refugee routes along the Burmese-Indian border. But does this mean I would remove those tracks if someone added to OpenStreetMap? This and other questions remain to be answered.

Look forward to trying it out in person next month, when Erica and I are in the UK (from June 17-24). Erica will be giving a talk on Map Kibera at my alta mater, University of Sussex.

With imagery, we’re trying to pin down the legal rights and restrictions on using imagery. We’re pretty sure we can derive vectors, just need absolutely clarity. There are probably restrictions on who can have access to the data .. this is commercial imagery after all. Not ideal, but we’ll see what’s possible. In amazing timing, the DigitalGlobe sales team is on a retreat this week, meaning we can’t get answers right away.

There are several promising contacts for help with place names. Takes a little time to bring people up to speed on the OSM tools; often best to be sitting at the same desk. Tomorrow, there are protests scheduled all over the US, and here in SF. I’m going out to try and meet some folks to contribute to the map. Feel free to do the same.

At the same time, we’ve collected a tremendous resource of existing geographic information and discussion on the wiki. I am really blown away by the quality and volume of data. Honestly, wow.

The European Commission offered to contribute their roads data. This is the data set that underlies UN-OCHA’s maps. However, the EC restricts use of this data to non-commercial, and doesn’t permit modification. I’m unsure about modification, since that’s exactly what OCHA does. But certainly non-commercial prevents use in OpenStreetMap. Commercial use is extremely unlikely in Gaza; if there was any commerce in Gaza, we should be happy. Anyhow, this data is now available non-commercial, non-modifiable. If anyone thinks this is useful, let me know, and I’ll arrange to host it somewhere.

Thanks everyone who’s emailed, blogged and tweeted in the past few days. Phenomenal response. Really, it doesn’t take much work individually for us to collectively do something awesome.

* The southwest has been traced pretty extensively. Incredible work from OpenStreetMap mappers.
* We’re looking into purchasing satellite imagery for the north or the entirety of Gaza. There’s actually B/W imagery available from yesterday! This may require a fund raiser, depending on the quote we get.
* JumpStart International has offered to host that imagery on the server, for tracing into OSM.
* We’re getting closer to bringing on Gazans living abroad, to add street names, place names, map features.
* AlertNet ran a story on the effort yesterday.

Thanks to all the blog posts and tweets to get the word out.

If you want to help, the OSM Gaza wiki is the place to start. It’s probably the most comprehensive listing on Gazan mapping generally. Or get in touch direct with me by email. Thanks.

Derinkuyu, ancient underground refuge city. via Warren Ellis”.

Excavated ant colony. Placement of fungal gardens and trash rooms (the bulbous nodes between air tunnels) creates heat gradients, resulting in air flow and CO2 emissions to the surface. via BoingBoing

wuo-uh, wuo-uih, wiou-uihh, whioucuhhh, hcoichr, AAIYhcAhcAIYcHoich, wherwlll, AIIchAYA, I!IAI!I

Wilfried asked a bunch of humans to transcribe this ape call. You should do it too, it’s totally fun. Then read the collected transcriptions.

So I did something about it. Finder didn’t turn up anything, so searched and found shapefiles from the state and the city. Processed. And uploaded to GeoCommons, and made a map of San Francisco Electoral Districts: City Supervisors, State Assembly and Senate, and Congressional. All in one map. Bravo GeoCommons!

I mean, it could be better of course. The path is not as smooth as it could be by a long way. I’ll complain below. And the result is a little hard to visually parse .. though that’s the nature of overlaying polygons. Anyway, by toggling layers, it’s easy enough to compare two sets of district boundaries, and see that the differences make no sense whatsoever. Where’s the revision history?!

Issues

Really the primary issue is how much can reasonably be done on the web. GeoCommons needs to address a wide non-professional audience, without the burden of GIS. What kind of operations can commonly be expected and incorporated into a simple workflow, and what is left off the 80/20 cliff.

When I first uploaded the assembly district boundaries to Finder, and viewed in the map, there was nada. Turned out the shapefiles are in a California State Plane projection, and Finder doesn’t notice or complain. There’s no way to submit a “.prj” file .. yet. So I had to do the transformation myself. Used OGR.


ogr2ogr assembly_districts_4326.shp assembly_districts.shp -f "ESRI Shapefile" -t_srs EPSG:4326


I then tried uploading again. It works! But dang slow. 80 detailed assembly district polygons in California. But I’m only interested in the two overlapping San Francisco. And I don’t want to simplify .. the precise boundary is crucial in this application. Couldn’t I just tell Maker that for my map, I’m only interested in features in this particular view? Seems simple enough to explain to the user, and generally useful. But not there yet.

Finally the styling. It’s maybe impossible to style four polygon layers in such a way that they’re all discernible in a single view.

GeoWeb!

But yea, GeoWeb! Here’s an interactive web map, with reusable data, without any programming required, just tools. We’re on the right path.

Tomorrow we fly to Nairobi. So much to write about, maybe someday I’ll catch up with life.

Burning Man camps and art in OpenStreetMap

Burning Man Garmin map files choose the latest

There’s a long chain and network of people and processes that got Black Rock City data into a fungible form, and a ton of cool imagery and tracking projects. I can talk about that next week, gotta pack.