Monthly Archives: October 2010

Technologies and Practice for the Prevention of Mass Atrocity Crimes

I’ve waited years for a conference like this: “Early Warning for Protection: Technologies and Practice for the Prevention of Mass Atrocity Crimes.”

This high-level conference combines my main areas of interest: conflict early warning, crisis mapping, civilian protection and technology. I’ll be giving a keynote presentation on “The Potential of New Technologies in Conflict Early Warning” at this conference next week, and I’m particularly looking forward to the panel that will follow, co-organized with my colleague Phoebe Wynn-Pope.

The conference will explore a number of issues.

  • What is the role of new technologies in conflict early warning and how do they interact with more traditional monitoring systems?
  • How can we harness, coordinate, and utilize the sometimes overwhelming amount of information available?
  • What systems and mechanisms need to be put in place to ensure effective early-warning is given?
  • How does the humanitarian sector work effectively with communities at risk once early-warning has been sounded?
  • How can a change in attitude and behavior at a policy level be brought about in a way that forestalls a descent to violence?

In preparing for the presentation, I started re-reading some papers I had written several years ago including this one from 2008: “Bridging Multiple Divides in Early Warning and Response: Upgrading the Role of Information and Communication Technology” (PDF). I will base my presentation in part on this paper and welcome any feedback readers may have. If you don’t have time to read a 25-page paper, here’s a short summary in bullet point format:

  • The field of conflict early warning has largely been monopolized by academics who are obsessed with forecasting conflict.
  • Operational conflict early warning systems are little more than glorified databases.
  • The conflict early warning community’s track-record in successfully predicting (let alone preventing) armed conflict is beyond dismal.
  • State-centric and external approaches to conflict early warning and rapid response have almost systematically failed.
  • The disaster early warning community have long advocated for a people-centered approach to early warning given the failures of top-down, institutional methods.
  • The disaster early warning community has been an early adopter of new technologies, particularly those engaged in public health.
  • The purpose of a people-centered approach is to empower individuals so they can mitigate the impact of a disaster on their livelihoods and/or to get out of harm’s way.
  • Preparedness and contingency planning are core to a people-centered approach since natural hazards like earthquakes can’t be easily predicted let alone stopped.
  • Given the dismal failure of conflict early warning systems, the conflict prevention community should make conflict preparedness and contingency planning a top priority.
  • Precedents for a people-centered approach to conflict early warning  exists in the fields of strategic nonviolent action and digital activism.
  • More importantly, communities that experienced conflict have developed sophisticated coping strategies to evade and survive.
  • Some of these communities already use technologies to survive.

I will expand on these points with several real-world examples and, more importantly, will combine these with what I have learned over the past two years, specifically in terms of crisis mapping, new technologies and civilian resistance. I’m excited to put all of my thoughts together for this conference, and I especially look forward to feedback from readers and conversing with participants.

 

Crowdsourcing the Angry Skies: The SKYWARN Volunteer Network

SKYWARN is a volunteer network of 290,000 trained storm spotters who provide localized weather reports to the US National Weather Service (NWS).  The concept was developed in the late 1960s and comprises a network volunteers who report “wind gusts, hail size and cloud formations that could signal a developing tornado” where they live.

According to Weather.gov, “the information provided by SKYWARN spotters, coupled with Doppler radar technology, improved satellite and other data, has enabled NWS to issue more timely and accurate warnings for tornadoes, severe thunderstorms and flash floods.”

This illustrates how crowdsourcing can be combined with “techsourcing” to provide better results.

Who Are SKYWARN volunteers?

Volunteers include police and fire personnel, dispatchers, EMS workers, public utility workers and other concerned private citizens. Individuals affiliated with hospitals, schools, churches, nursing homes or who have a responsibility for protecting others are also encouraged to become a spotter. NWS encourages anyone with an interest in public service and access to communication to join the SKYWARN program. (1)

Why Join SKYWARN?

There can be no finer reward than to know that their efforts have given communities the precious gift of time–seconds and minutes that can help save lives. (2)

How Are Volunteers Trained?

NWS has 122 local Weather Forecast Offices, each with a Warning Coordination Meteorologist, who is responsible for administering the SKYWARN program in their local area. Training is conducted at these local offices and covers:

  • Basics of thunderstorm development
  • Fundamentals of storm structure
  • Identifying potential severe weather features
  • Information to report
  • How to report information
  • Basic severe weather safety

Classes are free and typically two hours long. To find out when a SKYWARN class will be conducted in local your area, contact your local Warning Coordination Meteorologist. (3)

What else?

In some areas where Emergency Management programs do not provide storm weather reports, people have organized SKYWARN groups that work independent of a parent government agency and feed valuable information to NWS. While this provides the radar meteorologist with much needed input, the circuit is not complete if the information does not reach those who can activate sirens or local broadcast systems.  To this end, SKYWARN also distributes information from the National Weather Service. (4)

So What?

There has been much talk about the potential role of “Volunteer Technical Communities” in the context of disaster response. VTCs, as they are now called, came to the fore in the wake of the Haiti earthquake when their crisis mapping efforts helped the US Marine Corps and US Coast Guard save lives. VTC is the new buzzword, but technology-able volunteer communities have been around for decades. SKYWARN has been active for almost half-a-century.

As my colleagues and I continue to operationalize the Standby Volunteer Task Force (see this blog post and recent article on CNN), it behooves us to learn as much as possible from others who have set up volunteer networks in the past and in other sectors. The SKYWARN example shows how volunteer networks can interface with formal organizations in an effective manner.

The Spotter Network is a newer and less formal volunteer community that is not sanctioned or affiliated with the NWS or any other government agency. Nevertheless, “several National Weather Service employees and other officials have taken an interest in the capabilities [that this network] brings to them to integrate ground truth provided by spotters into their operational responsibilities. All at zero cost to them.”

The National Weather Service has responded positively to increasing public participation by launching the eSpotter, a system “developed to enhance and increase timely & accurate online spotter reporting and communications between spotters and their local weather forecast offices. The use of the system is currently available for trained spotters and emergency managers.”

Conclusion & Recommendations

  • Volunteer groups and government organizations can work together.
  • Volunteers networks include professionals as well as amateurs.
  • Training is an integral component of volunteer technical networks.
  • Government participation is key to leveraging volunteer groups.
  • Government can provide the infrastructure for collaboration.
  • Government reps should sit on the board of volunteer networks.
  • Generating unique data sets will get government attention. Fancy technology, bravado and media coverage won’t.

Analyzing Call Dynamics to Assess the Impact of Earthquakes

Earthquakes can cripple communication infrastructure and influence the number of voice calls relayed through cell phone towers. Data from cell phone traffic can thus be used as a proxy to infer the epicenter of an earthquake and possibly the needs of the disaster affected population. In this blog post, I summarize the findings from a recent study carried out by Microsoft Research and the Santa Fe Institute (SFI).

The study assesses the impact of the 5.9 magnitude earthquake near Lac Kivu in February 2008 on Rwandan call data to explore the possibility of inferring the epicenter and potential needs of affected communities. Cellular networks continually generate “Call Data Records (CDR) for billing and maintenance purposes” which can be used can be used to make inferences following a disaster. Since the geographic spread of cell phones and towers is not randomly distributed, the authors used methods to capture propagating uncertainties about their inferences from the data. This is important to prioritize the collection of new data.

The study is based on the following 3 assumptions:

1. Cell tower traffic deviates statistically from the normal patterns and trends in case of an unusual event.
2. Areas that suffer larger disruptions experience deviations in call volume that persist for a longer period of time.
3. Disruptions are overall inversely proportional to the distance from the center(s) of a catastrophe.

Based on these assumptions, the authors develop algorithms to detect earthquakes, predict their epicenter and infer opportunities for assistance. The results? Using call data to detect when in February 2008 the earthquake took place yields a highly accurate result. The same is true for predicting the epicenter. This means that call activity and cell phone towers can be used as a large-scale seismic system.

As for inferring hardest hit areas, the authors find that their “predicted model is far superior to the baseline and provides predictions that are significantly better for k = 3, 4 and 5″ where k represents the number of days post-earthquake. In sum, “the results highlight the promise of performing predictive analysis with existing telecommunications infrastructure.” The study is available on the Artificial Intelligence for Development (AI-D) website.

In the future, combining call traffic data with crowdsourced SMS data (see this study on Haiti text messages) could perhaps provide even more detailed information on near real-time impact and needs following a disaster. I’d be very interested to see this kind of study done on call/SMS data before, during and after a contested election or major armed conflict. Could patterns in call/SMS data in one country provide distinct early warning signatures for elections and conflict in other crises?

How Crowdsourced Data Can Predict Crisis Impact: Findings from Empirical Study on Haiti

One of the inherent concerns about crowdsourced crisis information is that the data is not statistically representative and hence “useless” for any serious kind of statistical analysis. But my colleague Christina Corbane and her team at the European Commission’s Joint Research Center (JRC) have come up with some interesting findings that prove otherwise. They used the reports mapped on the Ushahidi-Haiti platform to show that this crowdsourced  data can help predict the spatial distribution of structural damage in Port-au-Prince. The results were presented at this year’s Crisis Mapping Conference (ICCM 2010).

The data on structural damage was obtained using very high resolution aerial imagery. Some 600 experts from 23 different countries joined the World Bank-UNOSAT-JRC team to assess the damage based on this imagery. This massive effort took two months to complete. In contrast, the crowdsourced reports on Ushahidi-Haiti were mapped in near-real time and could “hence  represent an invaluable early indicator on the distribution and on the intensity of building damage.”

Corbane and her co-authors “focused on the area of Port-au-Prince (approximately 9 by 9 km) where a total of 1,645 messages have been reported and 161,281 individual buildings have been identified, each classified into one of the 5 different damage grades.” Since the focus of the study is the relationship between crowdsourced reports and the intensity of structural damage, only grades 4 and 5 (structures beyond repair) were taken into account. The result is a bivariate point pattern consisting of two variables: 1,645 crowdsourced reports and 33,800 damaged buildings (grades 4 and 5 combined).

The above graphic simply serves as an illustrative example of the possible relationships between simulated distributions of SMS and damaged buildings. The two figures below represent the actual spatial distribution of crowdsourced reports and damaged buildings according to the data. The figures show that both crowdsourced data and damage patterns are clustered even though the latter is more pronounced. This suggests that some kind of correlation exists between the two distributions.

Corbane and colleagues therefore used spatial point pattern process statistics to better understand and characterize the spatial structures of crowdsourced reports and building damage patterns. They used the Ripley’s K-function which is often considered “the most suitable and functional characteristic for analyzing point processes.” The results clearly demonstrate the existence of statistically significant correlation between the spatial patterns of crowdsourced data and building damages at “distances ranging between 1 and 3 to 4 km.”

The co-authors then used the marked Gibbs point process model to “derive the conditional intensity of building damage based on the pairwise interactions between SMS [crowdsourced reports] and building damages.” The resulting model was then used to compute the predicted damage intensity values, which is depicted below with the observed damage intensity.

The figures clearly show that the similarity between the patterns exhibited by the predictive model and the actual damage pattern is particularly strong. This visual inspection is confirmed by the computed correlation between the observed and predicted damage patterns shown below.

In sum, the results of this empirical study demonstrates the existence of a spatial dependence between crowdsourced data and damaged buildings. The results of the analysis also show how statistical interactions between the patterns of crowdsourced data and building damage can be used for modeling the intensity of structural damage to buildings.

These findings are rather stunning. Data collected using unbounded crowdsourcing (non-representative sampling) largely in the form of SMS from the disaster affected population in Port-au-Prince can predict, with surprisingly high accuracy and statistical significance, the location and extent of structural damage post-earthquake.

The World Bank-UNOSAT-JRC damage assessment took 600 experts 66 days to complete. The cost probably figured in the hundreds of millions of dollars. In contrast, Mission 4636 and Ushahidi-Haiti were both ad-hoc, volunteer-based projects and virtually all the crowdsourced reports used in the study were collected within 14 days of the earthquake (most within 10 days).

But what does this say about the quality/reliability of crowdsourced data? The authors don’t make this connection but I find the implications particularly interesting since the actual content of the 1,645 crowdsourced reports were not factored into the analysis, simply the GPS coordinates, i.e., the meta-data.

My Thoughts on Gladwell’s Article in The New Yorker, Part 2

The first part of my response to Gladwell’s article in The New Yorker explained why principles, strategies and tactics of civil resistance are important for the future of digital activism. In this second part, I address Gladwell’s arguments on high vs. low risk activism, weak vs. strong ties and hierarchies vs networks.

According to Stanford sociologist Doug McAdam, the civil rights movement represented “high-risk activism” which requires “strong-ties”. By strong-ties, McAdam refers to the bonds of friendship, family, relationships, etc. These social ties appear to be a necessary condition for recruiting and catalyzing a movement engaged in high-risk activism. “What mattered more was an applicant’s degree of personal connection to the civil-rights movement.” Indeed, you’re more likely to join a rally if your close friends are going. “One study of the Red Brigades, the Italian terrorist group of the nineteen-seventies, found that seventy per cent of recruits had at least one good friend already in the organization.”

In contrast, Gladwell argues that “the platforms of social media are built around weak ties.” The problem with evangelists of social media, according to him, is that they “believe a Facebook friend is the same as a real friend.” In addition, while “social networks are effective at increasing participation,” they only do so by “lessening the level of motivation that participation requires.”

Gladwell then adds the “networks versus hierarchies argument” to further his point. Strategic nonviolent action requires organization, planning and authority structures. Social media, on the other hand, “are not about this kind of hierarchical organization.” This is a “crucial distinction between traditional activism and its online variant,” says Gladwell.

Facebook and the like are tools for building networks, which are the opposite, in structure and character, of hierarchies. Unlike hierarchies, with their rules and procedures, networks aren’t controlled by a single central authority. Decisions are made through consensus, and the ties that bind people to the group are loose. This structure makes networks enormously resilient and adaptable in low-risk situations.

But it is simply a form of organizing which favors the weak-tie connections that give us access to information over the strong-tie connections that help us persevere in the face of danger. It shifts our energies from organizations that promote strategic and disciplined activity and toward those which promote resilience and adaptability. It makes it easier for activists to express themselves, and harder for that expression to have any impact.

I tried to summarize Gladwell’s arguments in the diagram below and would be interested in feedback. The red arrow represents high-risk activism and the green low-risk. As per his argument, high-risk activism requires both strong ties and high levels of organization.

Gladwell makes a compelling case and one that I largely agree with, but not completely. Would the four colleague students who instigated the first wave of protests in North Carolina during the Winter of 1960 have turned down the opportunity to use email, SMS, Facebook or Twitter? Would their use of social media tools have caused their movement to fail? Would the strong-ties these students shared be diluted as a result of also being friends on Facebook? I personally doubt it, they would still have shown up at Woolworth.

Gladwell is right to distinguish between high-risk and low-risk activism but this is a false dichotomy. Not everyone in society faces the same kinds of risks, nor do they face the same levels of risk all the time. Total war in the Clausewitzian sense only holds true for thought-experiments. Indeed, a recent study study found that, “The average percentage of area covered by civil war […] is approximately 48%, but the average amount of territory with repeated fighting is considerably smaller at 15%.”

So if communities face a range of risks that span from low to high, then one would want to leverage both strong-ties and weak-ties along with appropriate organizational forms, offline tactics and social media tools. This means that both networks and hierarchies are needed; and that neither organizational form need remain static over time since risks are not static. Indeed, an effective social movement needs organizations that promote strategic and disciplined activity and networks that promote resilience and adaptability. Both are absolutely key to the practice of strategic nonviolent action.

There’s no doubt that the civil rights movement represented high-risk activism. Does this mean that the same methods used in the 1960s would work for high-risk activism in a country like Egypt, North Korea or even in Cambodia during the genocide that killed an estimated 1.2 to 1.7 million people?  Can the US media of the 1960s really be compared to North Korean media? As my colleague David Faris noted in a recent email exchange on Gladwell’s article,

The initial sit-ins may have been launched by a small group of people impervious to the danger, but they grew to 70,000 not only because close friends were doing it, but also because people saw acquaintances protesting, and decided that the level of risk required to participate had fallen. This is important because in the U.S. in 1960 the media were willing to report on these events. This is not the case across the authoritarian world, where news relayed by text and Twitter may be the only reliable source of information apart from your immediate circle of friends. By relaying information about the preferences of your weak ties, social media provide individuals with more accurate pictures of the preference-sets of other members of their community.

New social media tools don’t dictate the organizational form of the movement, they simply create more options. So a hierarchical organization can very well use new media platforms to conduct their own highly centralized movement. It’s just like the Ushahidi platform, it is a tool, not a methodology. If a group of protesters don’t put any serious time into planning their campaigns, identifying key strategies and tactics, training, drafting contingency measures, fund raising, etc., then the presence of social media tools will not explain why their protests are ineffective. It would be too easy of an excuse.

Ushahidi is only 10% of the solution

Here’s a graphic designed by my colleague Chris Blow that shows why technology is at most 10% of the solution (the context is Ushahidi but the principle applies more broadly). If a movement doesn’t take on “all the other stuff”, then it doesn’t matter whether members are part of a network, a centralized organization, have weak-ties or strong-ties, or whether they are in a high-risk or low-risk environment. They are unlikely to succeed.

My Thoughts on Gladwell’s Article in The New Yorker

Malcom Gladwell’s article “Small Change: Why the Revolution Will Not be Tweeted,” was forwarded to me by at least half-a-dozen colleagues after it was published just three days ago. I have purposefully not read other people’s responses to this piece so that I could write down my own observations before being swayed by those of others.

So what do I think? Finally, someone else is calling attention to the importance of  civil resistance (strategic nonviolent action) in the context of new digital technologies! This intersection is what I’m most excited about when it comes to the new tools of social media.

Gladwell uses the example of the civil rights movement, which in his own words was an example of “high-risk activism” and “also crucially, strategic activism: a challenge to the establishment mounted with precision and discipline.” Indeed, “the civil-rights movement was more like a military campaign than like a contagion.” Gladwell is spot on, strategic nonviolent action is nonviolent guerrilla warfare. If I’m not trained in civil resistance, then I can still use all the technology I want but the tools won’t necessarily make me more effective or make up for my lack of skills in nonviolent warfare.

But most tend to completely skip over the rich lessons learned from the long history of nonviolent action because they are more excited about the tools. As Gladwell notes, “Where activists were once defined by their causes, they are now defined by their tools.” But these tools were never used in the vast majority of protests in the history of the world. See this piece by the Global Post on “How to Run a Protest without Twitter.”

I specifically blogged about this issue two years ago in a post entitled: “Digital Resistance: Between Digital Activism and Civil Resistance.” Some excerpts:

The future of political activism in repressive environments belongs to those who mix and master both digital activism and civil resistance—digital resistance. Digital activism brings technical expertise to the table while civil resistance offers rich tactical and strategic competence.

At the same time, however, the practice of digital activism is surprisingly devoid of tactical and strategic know-how. In turn, the field of civil resistance lags far behind in its command of new information technologies for strategic nonviolent action.

In this blog post, I called attention to the work of Gene Sharp who is considered by many as one of the most influential scholars in the field of civil resistance. His book, Waging Nonviolent Struggle, is a must-read for anyone interested in strategic nonviolent action. I argue that digital activism needs  much stronger grounding in the tactics and strategies of nonviolent civil resistance. That is why I followed up with a second blog post in 2008 on “Gene Sharp, Civil Resistance and Technology.”

In The Politics of Nonviolent Action, Gene identifies 198 methods of nonviolent protest and persuasion. The majority of these can be amplified by modern communication technologies. What  follows is therefore only a subset of 12 tactics linked to applied examples of modern technologies. I very much welcome feedback on this initial list, as I’d like to formulate a more complete taxonomy of digital resistance and match the tactic-technologies with real-world examples from DigiActive’s website.

So when starting from the principles, strategies and tactics of civil resistance, I do think that the tools of social media can act as multiplier effect in a nonviolent campaign. Gladwell rightly likens the civil rights movement to a military campaign. And communication is central to the effectiveness of nonviolent campaigns. In fact, some of the most successful nonviolent campaigns detailed in numerous case studies turned on the ability to get accurate, timely information. The literature on military history also demonstrates that “success in counter-guerrilla operations almost invariably goes to the force which receives timely information.”

Effective civil resistance requires sound intelligence and strategic estimates. But Gladwell only dwells on the role of new technologies in the context of recruitment. He doesn’t consider the effect of new tools on information sharing and information cascades. And if Gladwell had the time to read more of McAdam’s work, he’d have come across other relevant causal mechanisms described in the literature that are relevant to the discussion.

I plan to follow up with a second post based on Gladwell’s piece to address his points on strong versus weak ties and hierarchies versus networks.

Crowdsourcing the Analysis of Satellite Imagery for Disaster Response

I recently got a call from a humanitarian colleague in the field who asked whether it would be possible to crowdsource the basic analysis of satellite imagery.  They wanted to know because their team was sitting on a pile of satellite imagery but did not have the time or  staff to go through the high-resolution pictures. They wanted to use the imagery to identify where IDPs were located in order to know where to send aid via helicopters.

My colleague’s question reminded me of the search for Steve Fossett, a famous adventurer who went missing in September 2007 after taking off from a small airport in Nevada in a small single-engine airplane. The area where Steve went missing is particularly rugged terrain. The search and rescue aircraft were not able to find any sign of wreckage. However, high-resolution satellite imagery from GeoEye enabled Amazon to produce a Help Find Steve Fossett website, allowing volunteers to search small sections of the available imagery.

“This is an approach to more rapidly search a large area of imagery using many eyeballs of people around the world. A similar technique was used to search for Jim Gray, a Microsoft scientist who went missing on his sailboat off the coast of California.”

Micro-tasking the analysis of satellite imagery has already been done.  So why not in the context of disaster response? One could add this feature to a platform like Crowdflower, which is already being used as a plugin to micro-task the processing of text messages from disaster affected areas. Instead of text, volunteers would see a small subsection of satellite imagery. They’d be asked whether they could see any evidence of individuals in the imagery and if so how many approximately they can make out. A simple 5-minute guide on how to identify people and approximate population size using satellite imagery could be put on YouTube for volunteers to watch before getting started.

Like any type of micro-tasking approach (a.k.a. mechanical turk service), one could triangulate answers to maintain some level of quality control. For example, only when 10 volunteers each tag an image as having individuals in it would the picture be processed as such. The same would apply to the population ranged estimated in a given image. This wouldn’t necessarily produce perfect results, but it would take the bulk of the load off the shoulders of humanitarian on the ground. It would act as a first filter.

Of course the obvious question that arises is security and privacy. There are several ways this could be addressed. First, images would be stripped of any GPS coordinates. Second, images would be sliced up in small bits to prevent easy recognition of the territory. Third, a volunteer would not be given contiguous slices so they couldn’t piece together more information from the satellite imagery. These measures won’t provide 100% security and privacy. The only way to achieve that would be to use bounded crowdsourcing, i.e., only have trusted individuals analyze the imagery.