Disaster Response in the Digital Age

When natural disasters strike, people from around the world are inspired to assist, but are often constrained by distance, time, and the skills required to help.  But, with greater availability of robust web-enabled Information and Communication Technology (ICT), these barriers to participation may be cracking.  Technology is creating opportunities for a concerned public to participate more directly.  Aspects of the relief efforts to the January 2010 earthquake in Haiti offers glimpses of how disaster response may be changing in this globalized Internet Age.  But, do we understand how these seemingly novel actions came to be?  Did they make a difference?

In the aftermath of the earthquake in Haiti, technology simplified the process of soliciting and donating funds.  By sending a text message from a mobile phone with the word "Haiti" to 90999, Americans could donate $10 to the Red Cross.  As news of this and similar initiatives spread virally through websites like Facebook and Twitter and was amplified by mass media, the Red Cross raised five million dollars in three days via texts and over twenty five million in less than two weeks.  This was an impressive feat, but it was the acceleration of a familiar process, and alone does not make the response to Haiti groundbreaking.  We still do not know the ramifications of this phenomenon.  Perhaps instant donations could shorten the public's attention span.  By promoting unique short codes, will organizations confuse donors and affect other fundraisers?  Going forward, focus should also be placed on accelerating the disbursement of funds compared with their collection.

Numerous activities that emerged simultaneously to this tech-powered fund raising effort are worth investigating.  Via various web-based social networks, groups of all shapes and sizes consisting of thousands of people connected online and in-person to work on a flurry of projects.  Some developed technology platforms, others edited online maps, and more gathered in cities in a style reminiscent of software code writing "hackathons.”  The impact made and attention garnered by the unscripted synchronization of projects sets apart the Haiti response.

U.S. Secretary of State, Hillary Clinton noted that communication and information networks "played a critical role on the ground" and were a manifestation of a broader phenomenon by saying "the spread of information networks is forming a new nervous system for our planet."  Speaking at the Where2.0 Conference almost three months after the quake, Schuyler Erle, a technologist from SimpleGEO, labeled the aggregated actions as "unprecedented" because for the first time ever, there was a set of conditions, where “from the comfort and safety of one's own home can people help save lives in a disaster zone."  In opening the presentation, researcher John Crowley of the STAR-TIDES network, boldly stated, "What we did in Haiti changed disaster relief forever."

But did it really?  If so–how?  Can web-based, wiki-style collaboration complement on-the-ground aid delivery?  What benefits and challenges emerge when a "here comes everybody" approach is applied to a disaster zone?

--- ~ ---

Within four hours of the earthquake, Dewey Snavely, a heavy rigger who operates machinery for the Fairfax County Search and Rescue Team, mobilized in Virginia and prepared to fly to Haiti.  After the first team departed that night, he continued to prep with a second team and left the next morning.  As a veteran of numerous deployments, the presence of Mr. Snavely and the Fairfax team alongside the UN, military units, and various Non-Governmental Organizations (NGOs) is no surprise.  Yet even for him, Haiti "broke the mold on everything" due to the scale of the devastation and the fact that most of their local emergency management contacts had perished or were injured.^[1]

Technology played a new role, too.  In Haiti, he noticed that, everyone had a cell phone.  Once cell phone service had been restored, Mr. Snavely said the Fairfax team pursued leads and initiated a few rescues from nothing but text messages, which was "a first for us.”

Still, the devastation created complicating challenges.  Upon the team's arrival there were not many street signs or maps.  Checking for online maps the day before the earthquake would have yielded little, yet during their deployment, thousands of map updates were made that could be downloaded to handheld GPS navigation systems.  Snavely said the units were invaluable on this deployment as the maps were very close to reality.

--- ~ ---

In London, Dan Karran heard about the quake in the evening.  Mr. Karran, Lead Programmer at Inteeka, a web development company and contributor to the OpenStreetMap community, was at a pub meeting with some colleagues including mapping professionals and hobbyists.  When news broke, they looked to see what they could do, as did other open mapping advocates around the world.  They started tracing out Haitian streets and landmarks using a freely accessible application at OpenStreetMap.org that lets one edit maps using satellite imagery.  However, according to Mr. Karran, the initial imagery was poor and little could be done.^[2]

Within a day, the satellite imagery company GeoEye and Google Earth released high resolution imagery taken that morning.  With this unprecedented release, soon followed by other state and corporate imagery, OpenStreetMappers like Mr. Karran could get started.  They were joined by other experts and novices alike who worked solo or started to meet at gatherings such as Crisis Camps.

Crisis Camps—loose networks of people looking for a way to assist and armed with whiteboards, wi-fi, and laptops—drew heavily upon the ad hoc, open, and participatory style of workshop-events known as BarCamps.  Although BarCamp participants typically skew toward open source software and technologists who are accustomed to this dynamic, user-generated "unconference" format, surprisingly, “a different set of people showed up” at the London Crisis Camp said Alan Jackson, one of the organizers.^[3]  Most were professional people who hadn't been to such an event before, he said.  There, and in other cities like New York and Washington, D.C., campers started mapping and developing other technical solutions.  From just about any web browser pointed to OpenStreetMaps.org, people could make edits online.  Maps were taking shape and being downloaded.

One of Mr. Snavely's colleagues from the Fairfax County Search and Rescue Team did just that.  In a post on the Open Street Map wiki, a responder using the handle Rocklandusa said they were "thrilled" to have the maps on their Garmin GPS units and he spread the word among the humanitarian aid community.

When the quake struck, Open Street Maps had only two or three roads mapped in Port-au-Prince.  "Now, it is one of the most complete maps ... pretty impressive," said Mr. Karran.

--- ~ ---

In Boston, two hours after the quake, Patrick Meier picked up the phone and called the technical lead of his software team.  Over the past two years, Patrick and his team have been developing a Web-based election-monitoring platform named Ushahidi that can receive and map information from SMS messages sent via a short code.  This information is displayed on online maps.  Ushahidi (meaning testimony in Swahili) was developed and deployed initially in Kenya during the 2008 election unrests.  Now, Patrick thought, the same model could be applied to assist in directing responders to survivors in Haiti.  Within the hour, a basic version of their Ushahidi platform came online.  In twelve hours, with the help of a community of students at Tufts and other Boston-area universities, a near real-time crisis map of Haiti took shape.

The Ushahidi site became the public face of a number of autonomous initiatives already in motion that were soon to be integrated.  Josh Nesbit of the FrontlineSMS NGO along with Katie Stanton and Katie Dowd, two officials from the U.S. State Department, sought to repurpose an SMS gateway that Ushahidi could use for emergency messaging.  Via Twitter, Mr. Nesbit connected to the IP manager of Digicel, one of the two Haitian mobile firms.  Soon, the 4636 short code launched — a number to which Haitians could send requests for help and receive information.  The backbone of this system was the Emergency Information Service (EIS) developed in 2009 by the technology non-profit InSTEDD whose founding donors include Google.org.

Another component being put in place were workflows that allowed distributed Creole speakers to help translate SMS messages.  This process is an example of what is often called crowdsourcing, where information and contributions are requested from a distributed, web-connected group of people.

Within hours following the quake, the Thomas Reuters Foundation called upon InSTEDD to deploy the EIS for the first time in Haiti.  With EIS stood up after four days, the full integrated set of services was operational.  It enabled Haitians to receive information updates and send requests for help via '4636' where it would be translated and plotted by crowdsourced volunteers, and then could be acted upon by responders from the UN, Red Cross, and military responders.  In less than two weeks, this socio-technical system allowed participants to process 10,000 text messages.

Craig Clark of the U.S. Marines was one of many U.S. government representatives to lavish praise on the effort.  Speaking of Ushahidi, he said it saved lives and was used by the Marines "every second of the day" and was making the biggest difference of anything he had seen in the open source world.

--- ~ ---

These cases blur the boundaries among public/private, civil/military, corporate/non-profit, professional/amateur response initiatives.  Technology, represented foremost by mobile and web-based communications, afforded people the opportunity to participate by making the barrier to entry so low.  People who wanted to do more than donate money, but less than travel to the site, could take on small tasks, like mapping, that are easy to learn and carry out.  Supported by a new technical and social architecture, these efforts were meaningful to them personally, and their individual participation in aggregate amounted to a significant body of work.

In their presentation at the Where2.0 Conference, Crowley and Erle looked to Yochai Benkler to provide some explanations.  In his book, The Wealth of Networks, Benkler writes about how a networked information economy is changing many facets of society.  He defines this era as one where "important cooperative and coordinate action carried out through radically distributed, nonmarket mechanisms that do not depend on proprietary strategies" compared with the industrial information economy era.   While Benkler found evidence of this  "commons-based peer production"  in the development of open source software such as Linux,  the same factors that enable software development have, to a certain extent, bled over to complex crisis response.  As more and more people have access to computing tools, they are more able to contribute to information and cultural production –so why not contribute to the information tasks of disaster relief as well?  Benkler wrote that a networked environment makes possible “a new modality of organizing production” without relying on market signals or managerial commands.  It seems Haiti could be the first large manifestation of how this "social production" is altering another aspect of society.

While Benkler provides a framework for understanding how this phenomenon has come to be, it should be noted that all the initiatives that seemingly appeared overnight actually resulted from long development periods, not by "magic", said Ivan Labra, a researcher at the Naval Postgraduate School's (NPS) Collaboratory for the Study of Generative Networks.^[4]  He said efforts like OpenStreetMap were seeded by U.S. President Bill Clinton's decision to open up Global Positioning System (GPS) technology in 1996.  This directive made GPS open to civilian users and spawned a wave of new GIS and mapping activity.  Also, initiatives like Crisis Camps derived much from the behaviors established by open source software communities and BarCamp culture.  Mr. Labra cautioned that there is no short cut to crowdsourcing or to develop platforms like Ushahidi.  "These are things that are emergent only because a lot of work was done beforehand."

Technology enables innovation but it also spawns new challenges.  As more and more people converge and converse, the complexity of the environment grows.  What information is new and relevant?  How does valuable insight rise above the din of the blogosphere echo chamber?  Do discussions about a tragedy drown out the calls for help from victims?

In an insightful report on the use of information communication technologies in Haiti, authors Sanjana Hattotuwa and Daniel Stauffacher, from the ICT for peace foundation (ICT4Peace) noted that some critical challenges arose despite these innovations.  For example, multiple sites and systems for registering people created "islands and silos" of information that lead to "unnecessary duplication, fragmentation and significant frustration.”  As curators of data sources on the ICT4Peace wiki on Haiti, they found "myriad of data sources and proprietary formats ... suggests an increase in data fragmentation and lock-in” where only certain people and programs could access data. Beyond official sources, even the well-meaning can create an enormous amount of noise across social networks and social media sites by forwarding on conversations and re-tweeting messages that drown out new and useful information.

As much as disaster relief seems to have changed, many things didn't.  Traditional institutions still matter, especially the government of the affected populace.  Personal relationships, state powers, and entrenched response organizations remain key factors to response.  Organizations like the UN will likely be the primary enablers to integrate upstart technologists into response mechanisms.

Perhaps the characteristics of the crisis enlarged the role played by new actors.  The proximity to the United States and the scale of devastation in Haiti coupled with an already crippled government left such a large vacuum that more social entrepreneurs could participate.  This, in turn, attracted more media attention.  When an earthquake struck Chile about two weeks later, many sought to apply lessons from Haiti, but with less devastation and more infrastructure and functioning institutions in place, established organizations reacted better.  Less was needed and requested.  Therefore, there was less to discuss and fewer stories could be broken about revolutions in response efforts.  Does this signify that crowdsourced responses only work in specific conditions?  If so, what are the factors?

Despite the attention received by new technology and the role of the Internet, they are not a panacea.  In a vacuum, technology is useless.  People need to develop solutions and apply them.  Examples from Haiti showed how heavily the technologies depended on personal relationships and open social pathways to deploy and integrate in the right context. This is a situation where information technology like Twitter made social connections more quickly discoverable and enabled wider broadcast of requests.  Ever greater transparency has a trade off in the level of privacy, though, which needs to be recognized, as some push to move more communication flows online.

Both strong and weak ties among people are critical for information sharing as well as technology developments to be relevant.  While much effort was spent developing solutions at various Crisis Camps, they still require pathways through which to channel their work.  Alan Jackson explained that the London camp leveraged pre-existing personal relationships with contacts at the UN Office for the Coordination of Humanitarian Affairs (UNOCHA) for requirements and validation.  In his opinion, trust was the "key ingredient" in delivering useful solutions that could be deployed by those with a presence on the ground.  Camp participants form a loose collaborative network, but are not field experts.  Nonetheless, they can help as a scalable support team that can develop code, conduct usability reviews, and provide professional services on behalf of the larger effort.  Jackson said he thinks camp participants should be humble as to what they know and don't know about disaster response, vice striving to create the next disruptive technology.

Big organizations themselves still matter as well.  Whether the UN Cluster System was successful or not is being debated, but the lead coordination role played by the UN is certain.  They have a mandate and the organization to do so.  Emergent groups like the London Crisis Camp is directing its support to UNOCHA, while the ICT4Peace organization recommends continued support for the Chief Information Technology Officer of the UN Secretariat and other UN organizations, for example.

Additionally, state actors remain relevant as they bring to bear diplomatic clout and large amounts of money and manpower.  State Department officials were instrumental in persuading the second Haitian phone company to adopt the 4636 short code, for example.  Overall, numerous countries contributed funds and sent large response teams to assist.  The U.S. government is rare in that it could muster military and civilian assets at a price tag in excess of a billion dollars.

Therefore, it seems that these new responders are hardly going to dismantle existing hierarchies or replace the traditional responders, nor should they try. If anything, entrepreneurs need to make inroads into those organizations and assist them to become more nimble in their actions and accountable for their spending.  Upstart technologists should also learn from experts with field experience too.  Did Haiti change disaster response forever?  While it does not represent a total break from the past methods, the impact will reverberate in the future.  For example, it will be harder for nations and corporations to withhold satellite imagery during a future crisis, but the distributed development and use of online maps is “only one piece of a larger puzzle,” said Mr. Jackson.  While Crisis Camps may not have transformed responses, they demonstrated potential ways that it could be done, which is exciting, he said.  He continued to say that greater participation may even signal a growing sense of global citizenship where people feel motivated, able, and responsible to physically do something when disaster strikes, even if it is a remote ICT intervention.

With many players looking to participate, one hopes there is increasing mutual recognition of skills and assets, but time and resources will have to be spent for organizations to get to know one another and start planning together.  As Mr. Labra of the NPS Collaboratory points out, the emergent actions in Haiti were actually the result of decisions and actions taken long before.

Also, ICT4Peace recommends actors focus on standards-based information capture and exchange and the development of information sharing strategies ahead of time.  Now is the time to open doors to communication and develop methods to share information better, as to not create more "islands and silos."  No organization should waste time trying to command and control emergent actors, but instead find ways to absorb the contributions that passionate communities offer.  In his remarks, STAR-TIDES’ Mr. Crowley aptly stated, "For crowd sourcing… you require a crowd.  You need passion.  You need relationships.  And those take time to build.”

--- ~ ---

Beyond the hype and excitement over technology's role, successful responses depend on countless factors including speed, persistence, and luck.  Ultimately, survival more often depends on the resilience of the people affected.  After working to free one girl for 27 hours, Mr. Snavely finally reached out to her in the hole carved in the rubble.  Through a translator she said.  "Don't leave me, I'm not going to die here."

She survived.

--- ~ ---

^[1]Dewey Snavely. Phone Interview. April 15, 2010.

^[2]Dan Karran. Phone Interview. March 23, 2010.

^[3]Alan Jackson. Phone Interview. April 8, 2010.

^[4]Ivan Labra. Phone Interview. May 11, 2010.