Mark Iliffe (@markiliffe) is a geographer/map geek working on mapping projects around the world. He leads Ramani Huria for the World Bank, is Neodemographic Research Fellow at the University of Nottingham after completing his PhD at the Horizon Institute, and a mentor for Geeks Without Bounds.
Mark was interviewed for GeoHipster by Ed Freyfogle.
Q: Suitably for a geohipster, your OpenStreetMap profile says “I own a motorbike and have a liking to randomly spend weekends finding out ‘what is over there’”. What have you found?
A: I think I wrote that around a decade ago while getting into OSM, while on a foreign exchange trip in Nancy, France! I found out a lot of things, from that time trying to take a 125cc Yamaha (a hideously small and underpowered motorcycle — think Chimpanzee riding a tricycle) around Europe was slow and cold to new friendships. Also, a career path in maps and a love of all things geospatial, via counting flamingos in Kenya…
Q: Everyone has to start somewhere, and for you I believe that was mapping toilets (or places toilets should be). Indeed I think we first met when you presented your sanitation hack project Taarifa at #geomob by squatting on the table to demonstrate proper squat toilet technique. Tell us about Taarifa.
A: Taarifa is/was a platform for improving public service delivery in emerging countries. It came out of the London Water Hackathon in 2011, basically as an idea that we could do more with the data that is being generated by the many humanitarian mapping projects that had been enabled by OSM at the time, such as Map Kibera, Ramani Tandale and Haiti Earthquake mapping. As a community open-source project, it showed the potential of how feedback loops between citizens and service providers could be used to fix water points or toilets. We used Ushahidi as a base, adding workflow for reports; we tried to push these back to their community, but the core developers had other objectives — fair enough. We as the Taarifa community though we had something special regardless, but it was a hack, it wasn’t planned to be deployed anywhere.
In January 2012 I was in a meeting with a colleague at the World Bank who’d head that Taarifa had been suggested to fill a need on monitoring the construction of schools in Uganda. He arranged a meeting with the project manager for me, went along, and a week later I was coding on the plane to Uganda to pilot Taarifa across 4 districts around the country. Ultimately, it ended up being scaled to all 111 districts at the request of the Ugandan Ministry of Local Government.
From this the Taarifa community started to grow, expanding the small core of developers. In 2013 we won the Sanitation Hackathon Challenge, then received $100K World Bank innovation award to set up Taarifa in the Iringa Region of Tanzania. Taarifa and collaborators on that project, SNV, Geeks Without Bounds and ITC Twente then went on to win a DFID Human Development Innovation Fund award of £400,000. Since then it’s gone in a different direction, away from a technical community focus to one that concentrates on building the local social fabric that is wholly embedded and ran locally in Tanzania.
I feel that this was Taarifa’s most important contribution — not one of technology, but one which convenes development agencies and coders to innovate a little. Now, the main developers of the code haven’t worked on the main codebase for over a year, but Taarifa’s ideas of creating feedback loops in emerging countries still move on, in its grants, but also have been absorbed into other projects too.
Q: Actually I think I’m wrong, even before Taarifa you were an intern at Cloudmade, the first company to try to make money using OpenStreetMap. Founded by Steve Coast (and others), the VC-funded business hired many of the “famous” names of early OSM, before eventually fizzling out and moving into a different field. What was it like? Any especially interesting memories? What sort of impression did that experience leave on you? Also, what’s your take on modern VC-funded OpenStreetMap companies like Mapbox?
A: Cloudmade was fantastic, learned a lot from each of the OSMers that worked there — from Steve Coast, Andy Allen, Nick Black, Matt Amos, and Shaun McDonald. At Cloudmad, I wrote a routing engine for OSM — now common tools like PgRouting weren’t really around — I tried to build pgRouting from source, wasted three days, so started from scratch. In hindsight, I should have persevered with pgRouting, got involved in developing the existing tool instead of starting from scratch.
As it was my first tech company to work at, they were based in Central London and I was broke. I had to stay with my uncle in Slough about 30 miles away. I used to work quite late and slept in the office floor a few times. Once Nick was in early and caught me stuffing my sleeping bag back into the bottom drawer of my desk. The advice was to probably go home a bit more — advice that I’ve used selectively since, but I don’t sleep on my office floor anymore!
The VC situation is always going to be complex. I wasn’t too surprised when Cloudmade eventually pivoted, and their ideas and creations such as the “Map Style Editor” and Leaflet.js live on regardless of the company. At SoTM in Girona I made the comment that OSM was going through puberty. On reflection, I think it was a crude but accurate way to describe our project at that time. We didn’t know what OSM would or could become. OSM didn’t know how to deal with companies like Cloudmade, and neither did the companies know how to deal with OSM; to a certain extent I think we’re still learning, but getting better. Though at the time, like teenagers having to deal with new hormones, emotions ran riot. This all without realising that in the same way OSM has changed the world, OSM also is changed by it — and this is a good thing. Gary Gale has also mused extensively on this.
Now with the generation of companies after — CartoDB, Mapbox etc. — I think that they are much more perceptive to supporting and evolving the OSM ecosystem. Mapbox Humanitarian is one of them, but also their support for developing the ID Editor. In turn, the OSM community is growing as well, especially in the humanitarian space, with Humanitarian OpenStreetMap Team (HOT) supporting numerous projects around the world and acting as a useful interface to OSM for global institutions.
Q: Did you ever think back then that OSM would get as big and as global as it has?
A: TL;DR: Yes.
Recently, I had a discussion with a friend in a very British National Mapping Agency about the nature of exploration. Explorers of old would crisscross the world charting new things, sometimes for their own pleasure, but mostly for economic gain. These people then formed the mapping agencies that data from OSM ‘competes’ with today.
By working with the numerous army of volunteers, OSM embodies the same exploratory spirit — whether mapping their communities, or supporting disaster relief efforts. But instead of the privileged few, it’s the many. Now OSM is making tools and gaining access to data that make it easier than ever before to contribute, whether map data or any other contribution to the community.
Q: Despite those humble beginnings I believe you are now Doctor Mark Iliffe, having very recently defended your PhD thesis in Geography at the University of Nottingham. Congrats! Nevertheless though, doesn’t fancy book lernin’ like that reduce your geohipster credibility? In the just-fucking-do-it neogeo age is a formal background in geography still relevant? Is it something you’d recommend to kids starting their geo careers?
A: Thanks! Doing a PhD was by far the worst thing I’ve ever done, and will ever probably do — to myself, friends, and family. But it wasn’t through book learning, I did it in the field. Most of the thesis itself was written at 36,000ft via Qatar/British Airways and not the library (nb. This was/is a stupid idea, do it in the library).
Hopefully the geohipster cred should still be strong, but I wouldn’t recommend a PhD to kids starting their careers. Bed in for a few years, work out what you want to do, get comfortable, and then see if a PhD is for you. When I started my PhD, I’d done a small amount of work with Map Kibera and other places, and knew I wanted to work in the humanitarian mapping space but full time jobs didn’t exist. Doing a PhD gave the space (and a bit of money) to do that. Now these jobs, organisations, and career paths exist. Five years ago they didn’t.
Q: Though you live in the UK, for the last few years you’ve been working a lot in Tanzania, most recently with the World Bank. A lot of the work has been about helping build the local community to map unmapped (but nevertheless heavily populated) areas like Tandale. Indeed this work was also the basis for your PhD thesis. Give us the details on what you’ve been working on, who you’ve been working with, and most of all what makes it hip?
A: Ramani Huria takes up a lot of my time… It’s a community mapping project, with the Government of Tanzania, universities, and civil society organisations, supported by the World Bank and Red Cross. Ramani Huria has mapped over 25 communities in Dar es Salaam, covering around 1.3 million people. Dar es Salaam suffers from quite severe flooding, partly due as Dar es Salaam is the fastest growing city in Africa with a population of over 5.5 million.
Ramani Huria is powered by a cadre of volunteers, pulling together 160+ university students, 100s community members to collect data on roads, water points, hospitals, and schools, among other attributes. One of the key maps are of the extent of flooding, this is being done by residents of flood prone communities sketching on maps. Now that these maps exist, flood mitigation strategies can be put in place by community leaders — this could either be through building new drains, or ensuring existing infrastructure is maintained. That’s the hip part of Ramani Huria, the local community is leading the mapping, with ourselves as the international community in support.
Q: Over the last years there has been a big push by HOT and Missing Maps to get volunteers remote mapping in less developed areas like Tanzania. Some OSMers view this as a bad thing, as they perceive that it can inhibit the growth of a local community. As someone who’s been “in the field”, what’s your take? Is remote mapping helpful or harmful?
A: The only accurate map of the world is the world itself. With the objective of mapping the world, let’s work on doing that as fast as possible. Then we can focus on using that map to improve our world. Remote mapping is critical for that — but how can we be smarter at doing it?
To make a map of flood extents, so much time and effort goes into its creation. But a lot of it is basic, for example digitising roads and buildings. This is time-consuming — it doesn’t matter who does it, but it has to be done. But the knowledge of flooding is only held by those communities, nowhere else. The faster you can do this, the faster these challenges can be mitigated. Remote mapping gives a valuable head-start.
In Ramani Huria, we run “Maptime” events for the emerging local OSM community at the Buni Innovation Hub — these events grow the local community. Personally, I think we should move towards optimising our mapping as much as possible — whether that’s through remote mapping or image recognition — but that may be a step too far for the time being. I’d love to see interfaces to digitise Mapillary street view data, it’s something we’ve collected a lot of over the past year. Can we start to digitise drains from Mapillary imagery in the same way Missing Maps uses satellite imagery?
Q: You’ve recently been in Dunkirk in the refugee camps with Mapfugees, what was it like?
A: Mapfugees is a project to help map the La Linière refugee camp around Dunkirk, France. Jorieke Vyncke and I met up in Dunkirk to discuss with the refugee’s council — made up of the refugees themselves — and the camp administrators to see how maps could help. The refugees themselves wished to have maps of the local area for safe passage in/out of the camp. The camp itself is surrounded by a motorway and a railway, making passage in and out quite dangerous. Other ‘Mapfugees’ volunteers worked with mapping the surrounding areas with the refugees, leading local amenities and safe routes were identified.
At the same time, the camp itself was mapped, providing an understanding of camp amenities, so services to the camp can be improved. This is very similar to my experience of community mapping elsewhere — the map is a good way of discussing what needs to be done and can empower people to make changes.
Q: As you no doubt know, here at GeoHipster we’re not scared to ask the real questions. So let’s get into it. On Twitter you’re not infrequently part of a raging debate — which is better: #geobeers or #geosambuca? How will we ever settle this?
A: #Geobeer now has my vote. I’m way too old for #geobuccas as the hangovers are getting worse!
Q: So what’s next Mark? I mean both for you personally now that you’ve crossed the PhD off the list and also for OSM in places like Africa and in organizations like the World Bank.
A: For me, in a few months I’m going to take a long holiday and work out what’s next. I’m open to suggestions on a postcard!
Looking back, OSM is just past a decade old and is still changing the world for the better. In OSM, projects like Ramani Huria, but also mapping projects in Indonesia and others are at the forefront of this, but more can be done. I believe that organisations like the UN and World Bank need to move away from projects to supporting a global geospatial ecosystem. This isn’t a technical problem, but a societal and policy based concern.
This doesn’t sound sexy and isn’t. But at the moment, there are over a billion people that live in extreme poverty. Maps show us where to direct our resources and improve the lives of people, the human and financial resources required to map our world will be immense, moving well past the hundreds of thousands of dollars and spent on mapping cities like Dar es Salaam and Jakarta. To build this, we need to work at a high policy level to really embed geo and maps at the core of the Global Development Agenda with the Sustainable Development Goals. Projects like UN GGIM are moving in that direction, but will need support from geohipsters to make it happen.
Maps and geo are crucial to resolve the problems our world faces, to solve this problem we should use our natural geohipster instincts… JFDI.
Q: Any closing thoughts for all the geohipsters out there?
A: Get out there — you never know where you’ll go.