Mapping, Geolocation and the Future of Scalable Disaster Response

On Jan. 12, 2010 an earthquake of catastrophic proportions struck Haiti’s capital, Port-au-Prince. Among the buildings that were leveled was a school. In spite of the roof caving into the classroom, some of the children survived and one of them managed to send an SMS message. Relief workers, however, were unable to find the location of the school. Volunteers in Boston with Ushahidi were able to locate the source of the text message and sent that information back to the relief workers, who rescued the children.

This rescue was possible only due to the use of disruptive, community-driven Web 2.0 technology by volunteer and technical communities (VTCs) working on disaster and conflict management. VTCs such as OpenStreetMap, CrisisMappers,Crisis Commons,Sahana and Ushahidi have contributed greatly to disaster management. VTCs have used SMS, social media and satellite imagery; built communities around humanitarian efforts; and created technology tools and wikis, using open source software, hardware and platforms, as well as free cloud based services in affected countries such as Haiti, Libya and Japan.

Despite their successes, it has not been an easy ride.

Guest author Tanya Gupta is an international development professional by day and blogger by night. Her day job is in the Corporate Finance unit of the World Bank. At night, she reflects about development, technology and her past life in academia, as she writes.

VTCs continue to face major challenges, such as language and coordination. Many disasters occur in countries that are not English speaking, while much of the volunteer community is Angolophone. Coordination can be a problem too. Established development organizations such as the UN have been dealing with crises for many years and have a rich knowledge base, but are also challenged by data silos, proprietary systems and bureaucracy. VTCs are more agile and technically adept, but can be uncoordinated.

Disaster Relief 2.0: The Future of Information Sharing in Humanitarian Emergencies, a report produced by the UN and partner organizations, examines these issues in detail. It identified a host of additional challenges facing VTCs

  1. The need to build a reputation for reliability, trust, professionalism
  2. Lack of resources
  3. The technical challenges of geolocation with partial information and verifying accuracy of reports
  4. Building local capacity to manage disaster and conflict situations

It is true that this presents privacy concerns. However this may be a bigger issue in the West. […] In countries where privacy is not a cultural norm or expectation, geolocation software installed on the cheapest phones could provide enormous help during disaster relief efforts.

Despite these challenges the VTCs will play an increasingly important role in the future disaster management, thanks a to a growing number of volunteers and the power of Web 2.0 technologies.

So, what’s next?


We may see better methods of locating people during an emergency. Perhaps a Foursquare type check-in, or even better, an automatic check-in technology, where you don’t have to press a button to enter where you are, could be included in low-cost cell phones.

It is true that this presents privacy concerns. However this may be a bigger issue in the West. For example, the Singapore constitution does not contain any explicit right to privacy. In countries where privacy is not a cultural norm or expectation, geolocation software installed on the cheapest phones could provide enormous help during disaster relief efforts.

The countries with the highest number of people affected by disasters in 2010 include China, Pakistan and Thailand. These are countries where privacy protections are low, and where privacy is not a strong cultural value. They also score low on “individualism” in a framework developed by Greet Hofstede as a way to evaluate a country’s culture.

If we are to postulate that a lower score in individualism for a country also indicates that its people place a low importance on privacy, then it seems plausible that some disaster-prone countries could implement geolocation on cell phones without violating societal norms and save thousands, if not millions of lives.


A recent trend in VTC disaster management has been to use social media data as a layer on crisis maps. For example, a Hypercities project maps live Twitter messages on a map of Egypt, showing the location and picture of the Twitterer. This is helpful but some of the messages are clearly not relevant to crisis mapping. The challenge in using social media as a crisis map layer is that the data is huge, chaotic, free, and collectively good, but individually unreliable. To improve the social media data for inclusion on crisis maps, we need to focus on quality and relevance.

To determine quality in a Twitter stream, we can assume that source-quality equals information-quality. To identify quality sources for a given topic, we could use Twitter sources via curated lists from Listorious or established news media outlets and non-profits like Ushahidi. For instance Listorious has a list of reputable sources for the Haiti Earthquake curated by The New York Times.

To further refine the source-quality measure, we could also look at the number of followers of sources and the number of retweets that contain a relevant hashtag. For example, a tweet containing #civ2010 #IvoryCoast #civsocial #ict4d about the Abidjan crowd-sourced crisis map: was retweeted extensively in April.

We can assume that when we get higher quality tweets, the tweets are more relevant. Once the parameters for source and content quality are set, a program could read the Twitter stream and filter the quality tweets based on the selected parameters. In addition to improving the quality and relevance of the social media layer of crisis maps, perhaps we could also focus on improving the quality of the sources of the crisis map, through crowd sourcing methods such as incorporating Google’s +1 or a like-type function on information contained in the collaborative disaster maps.

Finally, improved matching between people needing assistance in a disaster and those who can help would add value. A service could be set up to match people affected by natural disaster with those who have the funds, goods, time or know-how to assist them. For example, Kiva partners entrepreneurs with lenders via existing microfinance institutions that facilitate the loans. matches American public school teachers who need classroom supplies with “microdonors.”

This kind of a matching service could be set up for organizations, groups, individuals and families affected by natural disasters. Some of the elements that would include skills, available time, specialty, needs (goods and services) and urgency.

The future of Web 2.0, social media and their applications are as unpredictable as the people they connect. But from what we have seen and and what we can reasonably postulate,  it is clear that these technologies have a profound positive impact in disaster management. I am sure the best is yet to come.

Photo by connor212

Facebook Comments