Having multiple resilient and trustworthy communication networks is a key requirement to be able to assess and react to a crisis, and technological developments offer new exciting opportunities to improve situational awareness, and document crisis information that would otherwise fall behind the horizon of the press and other conventional channels.
During an emergency, established communication channels may fail. One of the most effective channels to rapidly gather information is to tap into the only resource that is always present: the general public.
Crowdsourcing, the practice of aggregating information from a crowd, offers fast access to a great breadth of knowledge, since any person can provide intelligence. Beyond offering first-hand information that can be used to monitor the data from official communication channels or the press, crowdsourcing also has the additional benefit of empowering victims and responders, enabling them to help themselves and others. The creation of a crowdsourcing system bestows a direct, positive feedback on the field.
Since the public may be very diverse in terms of technical skills and access to technology, the most effective way to crowdsource information is to open multiple channels, and allow people to send reports using easy-to-use technologies they are already familiar with. All the gathered information needs to be standardized and displayed in a simple yet effective way to create a bigger picture and give immediate value to the citizens’ efforts.
The free, open-source Ushahidi engine is one such resource, allowing anyone to gather distributed data via SMS, email or web, and visualize it on a map or timeline. Ushahidi, which means “testimony” in Swahili, was initially developed to give Kenyan citizens the ability to map reports of violence after the post-election fallout at the beginning of 2008. This catalyst for a continued effort to develop a user-driven system that could become a tool for anyone around the world. Now the platform can be customized for different locales and needs, and it has already been used to track a variety of crises on a global (human trafficking, swine flu occurrences), regional (war in Gaza, stockouts of medical supplies at pharmacies in West Africa) and national scale (xenophobic attacks in South Africa, election monitoring in India, Afghanistan, and Mexico).
The experience gathered during these deployments has highlighted the need to assess the quality of the collected data. Some reports may be first hand, some may be rumors, and some may even be false. To assess the quality of information in the first the hours of a crisis, the Ushahidi team decided to expose the gathered “swift river” of information to a combined system: advanced algorithms automatically categorize the incoming reports, and, just like Wikipedia, a crowd of self-appointed editors can filter or edit the entries. This new project, dubbed “Swift River,” brings order to information coming from any source, whether it be crowdsourced – like through Ushahidi, Twitter, Flickr or YouTube – or institutionalized – the press, organizations, etc. The critical strength is the automatic categorization, as it clusters the reports according to the type and location of the events, allowing people to measure the trustworthiness of official and unofficial sources and triangulate the evidence to expose the probability that an incoming report is true or false.
In an emergency situation, this feature is fundamental, as it avoids wasteful management of resources while helping institutions and people to better navigate a difficult environment. Obviously, channeling the efforts of multiple players across the whole area affected by a crisis can multiply the effectiveness of disaster relief strategies. Moreover, systems like Ushahidi and Swift River can foster an increased level of accountability and transparency, since these free and easy-to-use systems can be deployed by anyone not just by resource-rich organizations.
Non-democratic countries may find it impossible to block spontaneous movements created by exasperated citizens, and the visualizations are so immediate that they can be understood even by non-experts. Even “hushed” emergencies can find a voice and the necessary coordination to tackle the issue, as users may share the evidence with their network, and create a movement.
In any environment, the ability to identify unreliable sources and expose truth-benders may trigger change. Thus, crowdsourced, automatically categorized data can empower people not just to collect their voices together and give a firmer ground for emergency response, but also become the cornerstone for social transformations, as they can identify and address crises that are not documented in official channels.