Self-Organized Crisis Response to #BostonMarathon Attack

I’m going to keep this blog post technical because the emotions from yesterday’s events are still too difficult to deal with. Within an hour of the bombs going off, I received several emails asking me to comment on the use of social media in Boston and how it differed to the digital humanitarian response efforts I am typically engaged in. So here are just a few notes, nothing too polished, but some initial reactions.

I Stand with Boston

Once again, we saw the outpouring of operational support from the “Crowd” with over two thousand people in the Boston area volunteering to take people in if they needed help, and this within 60 minutes of the attack. This was coordinated via a Google Spreadsheet & Google Form. This is not the first time that these web-based solutions were used for disaster response. For example, Google Spreadsheets was used to coordinate grassroots response efforts during the major Philippine floods in 2012.

We’re not all affected the same way during a crisis and those of us who are less affected almost always look for ways to help. Unlike the era of television broadcasting, the crowd can now become an operational actor in disaster response. To be sure, paid disaster response professionals cannot be everywhere at the same time, but the crowd is always there. This explains I have look called for a “ for disaster response” to match local needs with local resources. So while I received numerous pings on Twitter, Skype and email about launching a crisis map for Boston, I am skeptical that doing so would have added much value.

What was/is needed is real-time filtering of social media content and matching of local needs (information and material needs) with local resources. There are two complementary ways to do this: human computing (e.g., crowdsourcing, microtasking, etc) and machine computing (natural language processing, machine learning, etc), which is why my team and I at QCRI are working on developing these solutions.

Other observations from the response to yesterday’s tragedy:

  • Boston Police made active use of their Twitter account to inform and advise. They also asked other Twitter users to spread their request for everyone to leave the city center area. The police and other emergency services also actively crowdsourced photographs and video footage to begin their criminal investigations. There was such heavy multimedia social media activity in the area that one could no doubt develop a Photosynth rendering of the scene.
  • There were calls for residents to unlock their Wifi networks to enable people in the streets to get access to the Internet. This was especially important after the cellphone network was taken offline for security reasons. To be sure, access to information is equally important as access to water, food, shelter, etc, during a crisis.

I’d welcome any other observation from readers, e.g., similarities and differences between the use of technologies for domestic emergency management versus international humanitarian efforts. I would also be interested to hear thoughts about how the two could be integrated or at the very least learn from each other.


18 responses to “Self-Organized Crisis Response to #BostonMarathon Attack

  1. Pingback: 3 Observations: Social Media and the Boston PD #BostonMarathon | idisaster 2.0

  2. Pingback: #SMEM Crisis Response – The #bostonmarathon Bombing | buridansblog

  3. Your comment that the cell network was taken down has been denied by all major providers. Outages were due to heavy usage and NOT to MSP or anyone else requesting that providers take down their networks.

  4. The call for people to unlock their wifi was originated by Willow Brugh I believe as part of the digital volunteer response by Geeks Without Bounds in collaboration with multiple other teams and disciplines that Sara Farmer and others brought together to help build a public resource. It was a simple and effective solution. The team built a Boston Marathon Data page that was posted to the website, as an embedded iframe that updated live throughout the response phase, as well as a non-iframe mobile version updating hourly. See: Resources were shared widely on sm platforms including Facebook and Twitter as well as by SMS on a 1:1 basis with various Bostonians with whom the team members were connected. Much of this type of sm amplification may happen in the background behind firewalls and therefore difficult to validate after the fact, but the Willow wifi you cite is one example of the kind of reach that is possible for digivols to achieve by leveraging their own networks for the public benefit in an impacted area. I don’t see why any of this would be any different anywhere in the world, provided that the connections are there to leverage in the first place by a few local tech savvy volunteers. For me, it’s all about communication coordination and the will to break down silos, not fancy apps or tech. My cents.

    • Thanks Joanna, yes, I know that Willow and Sara were involved, and it is not the first time there has been a call for open wifi during disasters.

      “I don’t see why any of this would be any different anywhere in the world, provided that the connections are there to leverage in the first place by a few local tech savvy volunteers. For me, it’s all about communication coordination and the will to break down silos, not fancy apps or tech. My cents.”

      Thanks, I hardly think it is an either or. But processing/matching Big (Crisis) Data manually is a fantasy. Plus there’s absolutely nothing fancy about the apps or tech I’m writing about. This stuff has been around for a decade at least.

      • Perhaps I misspoke. I’m a fan of big data solutions and completely agree these are very much needed to manage volume and resource allocation to do the greatest good for the greatest number. No disagreement there, but my sense is that big data solutions currently do not have the ability (nimbleness) to respond with actionable information timely, meaning in the first few minutes of a sudden event of this type. A small group of trained volunteers using well honed manual processing and local knowledge can mobilize in literally minutes, as evidenced by the aforementioned GWOB effort. I truly believe the folks are needed to fill the gap until those with more advanced training and big data resources arrive. In the perfect world, these groups would not only be well connected in the impacted area, they would also have a pre-existing affiliation with the local agency that owns the event, so they could support a specific mission (VOST) and operate under the local Incident Commander. In the absence of that, as was the case of Boston, then to my way of thinking, communication coordination and leveraging those local connections becomes the number 1 priority, and I’m happy to report that significant progress is being made in that department in the world of VOST. Thanks for what you do.

  5. Pingback: Boston Marathon – How Can I Help? | Everyday Ambassador

  6. Speaking as a digital responder since the Iranian elections in 2009, I take pride in having worked with some of the finest colleagues in a number of venues and in response to numerous events. Most of these have been with well-oiled response teams working within each of their specific organizations.

    The opportunity of working with a coordinated team from mutiple organizations in response to the terrible bombing in Boston was unique. Willow and Sara created what became a fine collaboration of organizations and experienced digital responders, including both NGO’s and independent groups, local to Boston and Global.

    I write to applaud them and the leadership of the group, including Joanna, Pat Tressel, and Lindsay Oliver. And also to congratulate all the participants. It was an honor.

    Regarding the wifi callout, I would add that there are some standard items such as this that are essential statements in virtual response, #justincase they can be applied. For instance, I post first aid information during and following most events. Some people have commented that this information is either common knowledge or a waste of time, etc. On the contrary. Tens of thousands access this first aid information yearly. I know this thanks to data tracking.

    Working in the digital world we may never know who or how we help. But I know for a fact we do. Maybe not every suggestion or post applies in a specific instance, but basic information & good communication ideas are important to present for consideration at the very least.

    And, as always Patrick, thank you for your years of fine work. Invaluable.

  7. Pingback: Social Media for Good Roundup: Boston Marathon Bombings : Social Media for Good by @timolue

  8. Hi Patrick. I really appreciate your blog and perspective. I am wondering if there is a privacy issue with the display of full names, phone numbers, and email address in the screen shot you provide for this article? I think it provides great evidence, but I am wondering if those specific identifiers should be blurred.

    • Hi Jessica, thanks for your comment and for reading my blog. The screenshot was taken from a publicly available Google Doc which users voluntarily submitted their information in order publicly display said info to help others in need. So their seems to be informed consent. But now that the initiative is no longer needed, I agree that one should archive and close said Google Doc. I’ll remove the screenshot as soon as I’m back on my laptop. Thanks again for reading and for your comment. I welcome more feedback like this in the future.

  9. Hi Patrick, reading what Jessica wrote about the usage of screenshots….I am wondering if I can use the screenshots I have made for my Bachelor Thesis. What do you think? I believe this is publicly posted information and if I can reference the source I can use them. Is this right? Thank you for your help.

  10. Another observation, or issue that triggered my curiosity, was the self organisation in terms of “safe and well” messages. The American Red Cross had updated its platform (, and it was apparently overloaded – the server stalled few times. Yet most of Twitter messages were calling for using Google Person Finder – typically a non-intermediated service. Even further, Timo Luege suggests here the most efficient for that purpose was simply Facebook! (
    I come from an ICRC background where restoring family links is a traditional, essential activity that requires a heavy-process human layer. Sure, protection issues are different here and in conflict contexts, yet there is always a lot to learn from disasters. In that case, it seems the simpliest and most common the tools, the most used and thus the most efficient. So what could be needed in more comple cases is some kind of organizational API between specialist providers (like the Red Cross) and common tools (like Google’s)?

  11. Pingback: Thursday Morning Linkage » Duck of Minerva

  12. Pingback: The Best of Both Worlds: The Crowd and The Pro’s | Librarians Rule in a Crisis

  13. Pingback: Crowdsourcing Critical Thinking to Verify Social Media During Crises | iRevolution

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s