SXSWi 2010: Crowdsourcing crisis mapping with Ushahidi

Social media as a tool to save lives, document injustice

From South by Southwest Interactive 2010: Patrick Meier, director of Crisis Mapping and Strategic Partnerships at Ushahidi.com, demonstrates that open source crowdsourcing tools can make the difference between life and death.

Share:

Scott Laningham (scottla@us.ibm.com), developerWorks Podcast Editor, IBM developerWorks

Scott LaninghamScott Laningham, host of developerWorks podcasts, was previously editor of developerWorks newsletters. Prior to IBM, he was an award-winning reporter and director for news programming featured on Public Radio International, a freelance writer for the American Communications Foundation and CBS Radio, and a songwriter/musician.



19 March 2010

developerWorks: This is a developerWorks podcast from South by Southwest 2010 in Austin, Texas. Day four now. Scott Laningham here now with Todd Watson. And our guest is Patrick Meier, who is director of Crisis Mapping and Strategic Partnerships at Ushahidi.com. Welcome, Patrick. Nice to have you with us.

Meier: Thank you. Great to be here.

developerWorks: Tell us, can you give us a quick sense of what Ushahidi is?

Listen to this podcast.

The name of the organization is Ushahidi: Crowdsourcing Crisis Information.

Patrick Meier explains more on why he's at Ushahidi.

You can follow the mentioned in this article:

Meier: Be happy to, yes. Ushahidi is a free and open source platform that allows anybody to crowdsource — i.e., collect information in near real time and then to map that information in a dynamic, interactive, and interesting way.

Watson: So Patrick, this is Todd. You know, I don't know how this one escaped my radar, no pun intended, but as I've learned more about it I've become more and more fascinated by this technology, I think largely because this notion of crowdsourcing for real, you know, effective social purpose. And what I'd love for you to do is walk us back to 2007 and the Kenyan elections and how this came about.

Meier: Sure, be happy to. So I was raised in Kenya. I was visiting my parents there at the time. The elections turned pretty ugly quite quickly, got more and more violent, kind of spread around the country.

And some colleagues of mine basically saw this, Kenyans, and decided that something needed to be done. The media was not reporting on all the violence that was also being constrained somewhat by some political forces. The NGOs who were on the ground and had access to information were not exactly very forthright in sharing that information either. And so it was kind of frustrating to see that a lot of bad things were happening but none of this was being documented.

And if it's not mapped and documented, then for all intent and purposes, it never happened, right? If you can't prove that it actually took place if the first place. So Ory Okolloh, who is a prominent political blogger in Kenya, a Kenyan herself, was quite outright and critical of what was happening and had to flee the country because she was starting to get some death threats.

And that night she was able to escape with her baby girl and her husband to Johannesburg and she arrived that evening in Johannesburg and got online and brought up a blog post and said "all right, I made it out safely, one of the scariest moments of my life, but here we are. And why don't we, you know, can't we start using some mashups and Google Earth to start like documenting all this violence that's going undocumented."

And another colleague of ours, Erik Hersman, read the blog post ... who is a good friend, they all know each other in this tech Africa space ... and called up another Kenyan friend of his, David Kobia, and asked David whether he could put them together pretty quickly to see if that could be used.

And after a long weekend David came back with the first version of Ushahidi which is just a Web site with a simple map of Kenya, a Google Map of Kenya, and anybody could go on the Web site and submit an incident that they might have witnessed.

"Ushahidi," incidentally, is Swahili for "witness."

And so individuals could go online and say "just saw this riot happen in Kisumu, three people were beaten up" and then pinpoint it on a map where exactly they happen to have seen or heard about this.

The other innovative part to this was the fact that they also tied in SMS feature to the functionality with the map, so anybody could text a short code. The number was 6-0-0-7. Anybody in Kenya could text that number with their location and their testimony of what they saw or what they heard about and then that could be mapped.

So it was really bringing this crowdsourcing approach that, you know that we are familiar with perhaps in the technology community but bringing it to the humanitarian community. And even those technologies, even though those technologies were three years old at the time already, completely novel in the humanitarian community and something that generated a lot of excitement and interest. And that's how the story got started.

Watson: Wow. So I guess I'm curious as somebody who lives in the social media realm: Once you had the first Web site capability built and you added this SMS enablement, how did word get around to people who were maybe out in far-flown parts of Kenya to understand that they could even make these reports? I mean, there had to be a certain amount of growth of the idea for it to really blossom, right?

Meier: Sure. And so Ory Okolloh and Erik Hersman and David Kobia are very well sort of plugged into the Kenyan media, mainstream media scene, whether it's the TV, broadcast radio, print journalism in Kenya. So they were able to get the work out through their own networks.

And Ory, being the most prominent Kenyan blogger was getting 70,000 hits at one point a day on her blogs during the election violence. So it it took them writing a few blog posts, sending out a few emails through to the media and so on.

But also one thing that was interesting that I haven't actually mentioned it in a while is there were examples where local radio stations were printing out blog posts and reading them over local radio. And all of a sudden you've got a multiplier effect, right because obviously not everybody is going to be online, let alone be reading blogs in Kenya. But when you've got the radio stations that are starting to read blog posts over the radio, that word gets out even more.

So on Safaricom, largest Kenyan teleco, we worked with them for the short code. So they were able to also raise some publicity. So the word got out. It was where we as Ushahidi had a comparative advantage, being a Kenyan project and Kenyan initiative, we had many, many contacts there in the first place. And that's not the same with respect to Haiti or Chile or some of the programs or projects that happened after.

Watson: Yes, I actually wanted to maybe skip forward to some of these more recent natural disaster scenarios because that's obviously a very uniquely different scenario than an election and yet there's probably some consistent themes or threads to that.

What happened, for example, in Haiti and how was Ushahidi incorporated after that?

Meier: So the story is basically I was watching CNN at around 7PM, so about two hours after the earthquake had happened and I saw the news. I had actually some very good friends who were in Port-au-Prince at the time doing research. And I immediately called David Kobia, our tech lead at Ushahidi, who is based out of Atlanta and said "hey David, we really need to run with this."

So I come from the humanitarian side and I'm more and more involved in the technology side and I see a lot of opportunities sort of based on the experience that I've had working in the Sudan or Ethiopia in terms of how technology can be leveraged.

In any case, knowing that it was a 7.0 earthquake in a context like Haiti is basically a perfect storm. It's a very-very-bad-bad combination. And so immediately called David, probably around 7:30PM, and within an hour he had deployed a very basic, you know, version which was with very little customization which then we continued customizing literally for four weeks.

We deployed and we added the appropriate indicators, we added the text-messaging feature and everything else, and then reached out to a community of graduate and undergraduate students in Boston who then took the platform and did the near real-time crisis mapping virtually 24/7.

And because within about 12 hours we had the most comprehensive up-to-date information, actionable information on Haiti, that's how the UN found us, [US]SOUTHCOM found us, the Marine Corps, the Coast Guard, and so on, started then basically following the mapping that was happening ...

... and subscribing to our RSS feed to literally in the case of the Coast Guard and the Marine Corps to literally then have an alert come in say, of three individuals being trapped in a building and sending out the choppers as a response of that text message coming through and getting mapped on to Ushahidi near real time.

So we were intimately connected with the disaster-response operations which is not something I had necessarily anticipated, you know, when I decided to just act right away.

developerWorks: What kind of customization things do you anticipate from one experience to another? I mean, you said with minimal customization. What type of things do you have to do to be prepared for a Haiti over a Kenya?

Meier: So, some basic customizations that you can do that we don't really need any tech sort of background. And one of the things that we've done with the Ushahidi back end is modeled it along the lines of WordPress. We're big sort of WordPress friends; we think it's a quite intuitive user interface.

And so the idea is you download the platform and then you, you know, you come up with the title just like a banner kind of idea, right? But then you also choose which country. So there's a drop-down menu of all the countries.

You choose your country and then you have the ability to choose between Yahoo Maps, Google Maps, Open Street Maps ... a number of others as well depending on what sort of level of detail each of these maps provide and the kind of detail that you're looking for.

And then you would define the indicators. Are you looking for indicators on trapped individuals, food shortages, election violence, or what have you. Those can all be customized and you can upload whatever icons you want as well. And then you can define what email address you want. People can email you information.

So people can themselves enter information via Web form. Or they can text you information. They can also email you information.

In the case of Haiti, we had a dedicated haiti@ushahidi.com email address. We also put together, you could also ... we're also collecting information from Haiti ... from Twitter with hashtags /Haiti as well. So that's coming in.

We were also looking at, you can also aggregate tweets coming in and visualize them on the Ushahidi platform as well as Flickr and YouTube. So we were basically parsing Flickr and YouTube for any kind of references to Haiti disaster earthquake and pulling these in.

So what you have basically is an aggregating-type platform that takes this ecosystem of information communication technologies or this ecology of information and puts it in one place in a semi-structured way that can be used for multiple purposes.

developerWorks: Do you have a story about, you know, a use in the Haiti experience? I mean, is there something that you can tell about, you know, here's an example of the difference it made in such-and-such a situation?

Meier: Yes. I mean, there are a number. And in fact, it was hard to keep up in terms of being able to document everything. You have to remember that a lot of this we did during the first few weeks; we did this completely remotely.

developerWorks: And you weren't concerned with documenting it, you were concerned with helping people, obviously.

Meier: As were the disaster responders.

developerWorks: Right.

Meier: So even though we were asking them, please let us know, that is not something the responders are going to be focusing on when you're basically in a search-and-rescue operation phase.

developerWorks: Sure. That's not your priority.

Meier: And so and you know, we were getting a little heat as well just from the general public, like, prove it. Demonstrate the impact. And you know what was very helpful is we had the Marine Corps basically go on the record publicly and we blogged about this and we have the individual's entire email public where he's basically saying, you know, that he can't say enough about how valuable the Ushahidi Haiti platform is, that they are following it every hour of every day.

He gives an example of the 22nd Marine Expeditionary Unit that went out after one of the text messages that we mapped, literally, again, sort of sending out the choppers and saving three individuals. And then he goes on to say in the email saying "this is literally one of hundreds of examples that I can vouch for that your site has basically saved lives."

developerWorks: Is that cool or what? I mean, it's not the US military communication structure that was, I mean, sure it was significant, too, but that's a wonderful example of two different communities coming together to solve a problem fast.

Watson: Yes. So I would actually then be interested in learning if there is a potential negative or downside to this. I guess the question is broader in the sense, are there groups or individuals who could get this information and use it conversely? I mean, I think back to the election scenario.

In that situation, it sounds like the information was helpful and to a positive end. Could you envision scenarios where because it's open and anybody can get access, that they could use it for other more nefarious purposes?

Meier: That's a really important question. So there are a few things to say on this topic. One is to distinguish between what I call bounded crowdsourcing and unbounded crowdsourcing. Unbounded is sort of this very sort of pure crowdsourcing that we had maybe in ... which we had in Haiti and we had in Kenya.

But then there's also bounded crowdsourcing which is you basically have a trusted network of individuals reporting information and that information, the map itself and the data, does not have to be made public. It could be password protected.

Now, there's a trade-off between bounded crowdsourcing and unbounded crowdsourcing. If you're bounding it, you're not getting as much information, it's not as public. Maybe that prevents some synergies and catalyzing of synergies to happen.

But it's absolutely true, you know, the platform is being used by a number of different groups around the world in what we call non-permissive environments — i.e., authoritarian regime, countries under authoritarian rule.

So what we've been doing is working with a number of these partners to looking at how we make the platform even safer for individuals to use. I don't think you could ever get to a point where it's going to be a hundred percent necessarily, but there's certainly some measures you can take both on a technology standpoint as well as on an tactical standpoint that you can make yourself safer when you communicate in repressive environments.

One idea is to see how we can make people who submit information to Ushahidi, make that information non-traceable back to the individual who sent it, make it encrypted, make it anonymous, and so on. So something that we're very much continuing to work on and develop with folks who are experts in this kind of area, you know, encryption and security, data security, and so on. And that's thanks to a larger community that we have.

But it is an ongoing concern. And, you know, there's some groups who are looking to use this this year in a number of countries that are anything but democratic and who would definitely feel threatened by something like this and who would then go probably go after, you know, that particular group using the Ushahidi platform.

And so, another thing that I want to make sure is clear is we don't take the lead on the vast majority of these platforms. This is not our comparative advantage. At the end of the day, our comparative advantage is developing these innovative platforms, curating the code, and being demand driven.

And by that I mean we don't add a feature to the Ushahidi Web site or the platform because we think it's cool. Everything that's on there comes from a long wish list from current partners, perspective partners, other individuals in the humanitarian space, the human rights space, or media who say we would need a function like this. Just like they said, we need a way to report anonymously. So we're basically creating the code and adding these features. And that's our core advantage.

Humanitarian organizations, activist groups, political opposition parties in-country have a better idea of what the lay of the land is and what risks they are prepared to take and not take.

At no point are we ever involved in basically telling somebody you should use Ushahidi in Iran, right? It will be an Iranian group who comes to us and asks "hey, can we use Ushahidi" and we'll say "yes, it's free, download it, it's completely open source, best of luck. Let us know if there's some tech support that we can give." But we're not about to give any kind of strategic guidance or take the lead into how you deploy this.

We're happy to share lessons learned from other partners who have used it and said "we'll refer you to the folks who have used in it this country; they might actually be helpful because it's two repressive countries; you guys should talk." So we connect people and we share information. But we are in the back seat and that's how we prefer to be.

developerWorks: Great story, Patrick. Patrick Meier, Director of Crisis Mapping and Strategic Partnerships at Ushahidi.com, and Todd Watson here.

Again, South By Southwest 2010 Interactive, Day 4, developerWorks, more to come. Thanks for listening.

Resources

Comments

developerWorks: Sign in

Required fields are indicated with an asterisk (*).


Need an IBM ID?
Forgot your IBM ID?


Forgot your password?
Change your password

By clicking Submit, you agree to the developerWorks terms of use.

 


The first time you sign into developerWorks, a profile is created for you. Information in your profile (your name, country/region, and company name) is displayed to the public and will accompany any content you post, unless you opt to hide your company name. You may update your IBM account at any time.

All information submitted is secure.

Choose your display name



The first time you sign in to developerWorks, a profile is created for you, so you need to choose a display name. Your display name accompanies the content you post on developerWorks.

Please choose a display name between 3-31 characters. Your display name must be unique in the developerWorks community and should not be your email address for privacy reasons.

Required fields are indicated with an asterisk (*).

(Must be between 3 – 31 characters.)

By clicking Submit, you agree to the developerWorks terms of use.

 


All information submitted is secure.

Dig deeper into developerWorks


static.content.url=http://www.ibm.com/developerworks/js/artrating/
SITE_ID=1
Zone=
ArticleID=494044
ArticleTitle=SXSWi 2010: Crowdsourcing crisis mapping with Ushahidi
publish-date=03192010