Site Meter Haiti One Year On: Technology and the Future of Humanitarian Aid | UN Dispatch
Increase Font Size Decrease Font Size

Haiti One Year On: Technology and the Future of Humanitarian Aid

Ed Note: January 12 marks the one year anniversary of the devastating earthquake in Haiti. Over the next few days we will be running a series of posts taking a look at progress over the past year and what still needs to be done to help get Haiti back on its feet.

By John Crowley

The disaster response operations of 2010 taught the humanitarian community hard lessons in how to communicate more effectively both with each other and with disaster affected communities. In reports reflecting on the year since the Haiti earthquake, key members of our community (see Internews and the ICT4Peace Foundation) are pointing to the need to adapt practices in the humanitarian response system to new realities, including policies to protect the security of data about vulnerable populations.

Turning these recommendations into implementable practices is going to require input from many stakeholders, including humanitarians from large agencies as well as representatives of the volunteer technical community. The United Nations Foundation & Vodafone Foundation Technology Partnership and UN Office for the Coordination of Humanitarian Affairs (OCHA) have partnered with the Harvard Humanitarian Initiative to conduct these consultations and propose both a framework developing this long-term dialogue and defining how new tools and practices can be integrated into the realities of the work in the field.

Consultations to gather both information and recommendations toward a framework that facilitates coordination between international humanitarian institutions and volunteer technical communities are ongoing, and will be presented in a report to be released this March.  The report is intended to seed discussion between various stakeholders in emergency response and open a conversation through which practitioners and policy makers can discuss and seek solutions to technology-related emergency response challenges, including those seen in Haiti, Chile, Pakistan, and other emergency operations.

Based on dozens of initial interviews with staff from UN agencies, large NGOs, governments and the emerging community of humanitarian technologists and mappers, the Harvard team has distilled six of the core challenges, which are listed below.

The six findings so far include:

1.       Mobile technologies and satellite communications are bringing everyone—humanitarian organizations, international institutions, volunteer technical communities, and the affected populations—ever closer together. More often than not, victims of disasters and conflicts have cell phones and can communicate via SMS in real time. The headquarters of various agencies are ever more closely connected to the operations in the field. And thousands of volunteers who until recently would have sent donations are now contributing mapping and crowdsourcing services.

2.       As a result, information flows are accelerating, raising expectations around increasing the tempo of information management and coordination inemergency operations.

3.       At the same time, the methods for data and information exchange are moving from document-based systems to flows of structured data via web services. This movement from the narration of ongoing events in long stretches of unstructured prose to streams of data in short, semi-structured formats require humanitarian staff to perform double duty. They are simultaneously working within an existing system based on the exchange of situation reports while filtering and analyzing high volumes of short reports arriving via SMS and web services.

4.       Information Management in the humanitarian system is not tooled to compile, translate, and analyze the increased messaging from an affected population, the VTCs, or the demands of headquarters. For field staff who are working in difficult circumstances in technology-hostile environments, the sense of information overload is unprecedented and increasing.

5.       Field staff and their managers are saying that the best method for integrating non-traditional information flows with humanitarian information management practices is to link new data flows into existing workflows and shared data standards.  Adding new work flows will break the system.

6.       As a result, stakeholders are calling for an interface between the humanitarian system and its cluster coordination mechanism and the various new sources of information—from the disaster affected community and the volunteer/humanitarian technologists.

Add your voice to the discussion by using the comment field below or email us at newhumanitarian-at-gmail-dot-com.

John Crowley is leading a team of researchers at the Harvard Humanitarian Initiative working on this report.  Further information about the project is available here.  This report will build on findings from the 2009 report New Technologies in Emergencies and Conflicts: The Role of Information and Social Networks.


  • Guest

    Thanks for this post and these efforts.

    Bluntly speaking, there is nothing new or particularly insightful about these findings. We understand the role that mobile technologies and satellite communications are playing. We know that cell phones and SMS are important. This has already been reflected in the 2009 report and many other studies. We are also aware that volunteer communities are playing a more active role in disaster response hence initiatives like Crisis Commons and the Crisis Mappers. We know that information flows are accelerating and are aware of the expectations this can raise. We understand that complexities are associated with structured and unstructured information flows. We realize the limits of humanitarian information management systems and the need for basic standards. And finally, we all know about the need for an interface described in the findings.

    In sum, these findings provide nothing new to an already active community.

    • http://twitter.com/andrejverity Andrej Verity

      It is true that the “findings” listed here are not new, but the important element is that the report from OCHA, UNF, Vodafone and HHI will propose a framework on how the “traditional” humanitarian community and the VTC can work together (i.e. how to bring the “traditional” humanitarians together with the VTC in a constructive & productive way). And, THAT is what I have not seen before.

      It is easy for people to write that things need to change, that mobile is the wave of the future, that the humanitarian community needs to engage with the VTC, that the VTC needs to learn about the humanitarian community, etc, etc. But, it is much harder to devise a framework the enable such partnerships.

  • http://twitter.com/faisalchohan Faisal Chohan

    This is very nicely written. One of the biggest challenge is to embed new technologies and systems into existing echo system. The organizations are already under a-lot of burden in disaster and they do not have time and capacity to consider new technologies.

    Faisal Chohan
    Pakreport.org
    faisal (at) pakreport.org

  • http://twitter.com/faisalchohan Faisal Chohan

    It is important to work on the role of new technologies at times before disaster. Otherwise, there is high risk that too much effort is put in implementing and scaling these technologies.

  • http://twitter.com/spangledrongo Nigel Snoad

    Guest/Daniella,

    while it’s true that there’s already a bunch of reports and community knowledge on these points, I’d say the big step is that this work is an indication that the “traditional” humanitarian system is starting to understand these issues, and is actively asking for some things that “we” already know are needed. I’m most interested in how “we” decide what codes of conduct should apply (for example) and how the interfaces work. The volunteer Standby Task Force is one way that’s moving ahead (focussed around one tool) but if you assume the value of volunteer and technology communities then there have to be interfaces, and several. None of the earlier reports address these structural issues, they were often driven by case studies. And we don’t have a place for those either. Or evaluation frameworks. Or lots of other things.

    Nigel

  • http://www.google.com/crisisresponse/ Pablo Mayrgundter

    John, nice distillation! Here’s supporting evidence from the wall next to my desk at Google Crisis Response:

    http://picasaweb.google.com/pablo.mayrgundter/AirMe#5562136746128632818

    We asked people to write down short ideas/take-aways during a 2 day session on tech challenges in crisis response and saw an overwhelming consensus on data sharing. It seemed like it was a well known problem because it has been so resistant to solution :)

    My take on this is that it’s like death from a thousand (post-it note paper) cuts. No particular technical issue is a show-stopper, but together they constrain the solution space to personal computing tools paired with e-mail and relatively static websites and significant use of SMS as the most manual of all.

    What has been surprising to our group is how good a solution this is given the constraints. To us this has posed challenging questions, like, can we create powerful interfaces to our systems via email? Imagine multiple organizations cc’ing spreadsheet to a structured wiki, and anyone from twitter turks to field workers joining in with short messages. In some ways it seems like a step backwards, but maybe in preparation for a major advance.

    I really appreciate the work to distill this to its essentials; that’s the path forward for engineering.

    To paraphrase you, our tools need to enable access to increasing flows of structured data from a diversity of communication devices.

    Or from one of the post-its: “Crowdsourcing & Crowdfeeding”

  • Sanjay Rane

    This is a good post outlining current challenges in linking technology to disaster management.
    I partly agree to a comment by a Guest commentator to this post that the findings provide nothing new to an already active community. However, I also believe that refuting suggestions without contributing to solving the problem is also not good for community spirit.
    The world that we are living in is very complicated. If we try to invent a global solution to world problems then I believe we will fail again. My suggestion is that we should be looking at local solutions to global problems. This is where I agree to another commentator that we should be focusing more on building disaster resilient communities by empowering them with technology to reduce the impact of disasters. Map Kibera, in Kenya, is good example (best practice) of how it can be achieved.
    Empowering local youths in informal settlements around Nairobi to map their neighbourhoods not only helps building technical skills in using new technology but also engages them in fruitful activities. Large-scale flood mapping carried out by this empowered group during last flood season was never possible before. Maps they produced are now being used by development community to effectively plan humanitarian intervention projects.
    This project could lead us to exploring new ideas around community empowered information management.
    Sanjay Rane
    Information Management
    UN OCHA Kenya

  • Kashif

    I agree in principal to all the above findings. They are not new but true.
    Being an Information Management Specialist with UNICEF led clusters, the problems I am facing are (not in order of priority):
    a. Non coherent information flows (amongst stakeholders) and sometime no information flow.
    b. Non standardized datasets, reporting formats and types of datasets.
    c. Mostly no linkages of Attribute Data to Spatial Data
    d. Not enough capacity to handle information and its utilization for informed decision making.
    e. Data compatibility issues for joint planning
    f. Data storage and responsibility issues
    g. Ownership issues
    h. Language issues
    i. Pressing demands for real time information
    Although there is a long list of issues (in addition to mentioned above) which i think can be settled with in a reasonable timeframe. But time itself is the most critical issue in humanitarian response. I hope this framework addresses most of the issues if not all.

    • John Crowley

      Kashif, thanks for this thoughtful response. Your list of problems tracks with what we are hearing from other staff, both in the field and home offices. Would it be possible to talk with you about how you are currently confronting these challenges?

  • http://twitter.com/conneally paul conneally

    This is an extremely useful summary of the opportunities and challenges presented by new technologies and aid delivery. This project is really important for aid groups to follow and contribute to. At the IFRC.org we are about to announce a major new global partnership to enhance our early warning and preparedness (as well as disaster response) via use of SMS. You may also be interested in this: http://www.youtube.com/watch?v=Ouug-laNO8E&p=8011AD663CC06CF4

  • Jaro

    Hi, just a comment from the “other” side. I am one of the co-founders of standby taskforce (blog.standbytaskforce.com). For many of you and other people professionally active in this field on the side of agencies such as OCHA might be some of these findings already repeatedly identified. However, from a perspective of a tech and mapping volunteer and/or volunteer groups, it is critical to understand the challenges you guys are facing. And this understanding is critical for any constructive collaboration and effective support from crowdsourced volunteers to response operations. There are opinions that there is no space for amateur online volunteers in such serious situations such as disaster response. Well, let’s face it, there is no stepping back anyway, this process already started and more people will be seeking ways how to get engaged in such efforts (and actually, so many of them actually are professionals in relevant fields). We can try to ignore that and during emergencies simply treat them as annoying disorganized flies, or, we can improve their (our) understanding and awareness in between emergencies so they have better idea on how to perform the tasks in a way that it effectively addresses the needs of the hum community (both local and international responders). From this perspective, this report is really helpful, as well as the comments here in the discussion. And last comment @Nigel re Standby Taskforce: it is true that so far the SBTF was focused predominantly on one platform, but that is simply because the initial SBTF team’s skillsets were coming from an experience with working on this one platform during recent emergencies. But this is changing… we knew and announced from the beginning that other platforms and initiatives have to be brought on board in order to put together a functional framework and interface. And just these days we are having very fruitful conversations with different partners to make such partnership and collaboration come true soon. And besides that, we keep developing our internal protocols for better mobilization of volunteers, their organization during deployments, methods to maintain engagement of volunteers over time and improving their skillsets through trainings. Sorry, this post is long, and really not meant to be only as a promo on SBTF activities :-)

  • http://www.facebook.com/people/Josh-Knauer/650280378 Josh Knauer

    From my experience in designing, implementing and deploying large-scale, community-based collaborative data sharing systems, it seems to me that the “answer” (not that there could possibly be just one), would be to design and iteratively start building an open API (application programming interface) for humanitarian aid. I’m purposefully using a computer programming term for a non-computer problem to highlight the relative simplicity of the solution. An effective API requires two main components: inputs and outputs. Established aid agencies have existing systems (often non-technical) that need/expect specific types of inputs. If they can clearly define what those inputs should be, then the Humanitarian Aid API (HAAPI… just making up another acronym here for the fun of it) could be written so as to always provide outputs to the inputs that are needed.

    Once those outputs are clearly defined, then any group, whether ad hoc or professional, that wants to participate effectively in humanitarian aid efforts has clearly defined targets for what their systems should be outputting as inputs to HAAPI.

    To do this effectively, I would propose that a fairly significant and iterative design process needs to be started that would help clarify the requirements of aid agencies and start building early prototypes based on “low hanging fruit” or the easiest problems that can be solved early and quickly. The great thing about an effective API process is that you can build them in a very iterative process over time, which means you don’t have to design everything and get it all right the first time.

    I realize I’m a new voice to this specific discussion, but both I and my company, Rhiza Labs, have an extensive amount of experience building community-based collaborative data sharing tools. The problems that are described in this article are quite common in many community-based data systems and there are quite a few examples of solutions that others have done in other domains. You can find some white papers on our website about some of our experiences in this area: http://www.rhiza.com I don’t want to come across claiming that we have all of the answers, because I’m certain we don’t, but it might be a worthwhile exercise to learn from the experience of other non-UN/non-humanitarian aid communities.

    I’m expressing these opinions because I wish to help. If I can be of any assistance to this community, please don’t hesitate to contact me and let me know.

    Thanks,
    Josh Knauer, CEO, Rhiza Labs
    http://www.rhiza.com – 412-488-0600 – knauer@rhiza.com

Diplo Tweets