By Yoh Kawano and Hiroyuki Iseki
Reasons must vary for participants joining the corps of IT volunteers following a major global disaster. In the case of the Japan Earthquake of March 11, 2011, two Japanese nationals living and working in the US decided to immediately submit their names and expertise to a pool of applicants. The reason they decided to volunteer was perhaps the reason why they were not only selected, but chosen to lead a group of six GIS experts from around the globe. The location of the disaster assistance project was their very home country, and their helplessness towards the situation turned into a desperate plea to allow them to pour their hearts and souls into ways they could passionately contribute towards the relief effort from abroad.
On March 12th, the day after the earthquake, Yoh Kawano, who works for UCLA’s Office of Information Technology and lectures in the School of Public Affairs, googled “GIS volunteer” while watching the horrors of the disaster unfold on his home TV. Clicking on the first result that showed up, he quickly learned that an organization called GIS Corps (http://www.giscorps.org/) was accepting volunteers.
He hastily put together his resume, and submitted it. The very next day, Kawano was interviewed via skype by the co-founder of the organization, Shoreh Elhami. As luck may have it, he suggested to her that Hiroyuki Iseki, a very good Japanese friend and current Assistant Professor at the University of Maryland, would be a great addition to the team. Soon, both were selected to join four other GIS experts from around the world (overall, three applicants were based in the US; the other three were in Greece, Canada, and England), and by that afternoon, they began their volunteer mission, only 36 hours after they first contacted GIS Corps.
The “mission” began slowly. While it appeared as though GISCorps was in charge of recruiting so called GIS “experts” for certain missions, it was CrisisCommons (http://crisiscommons.org/) who was running this particular operation. They were initially given three materials to get started:
- a document with a mission statement,
- a wiki site, and
- a skype channel invitation.
The mission document was rather vague, a request to collect “relevant” spatial datasets according the guidelines of Common Operational Datasets. The wiki site only added to the confusion, as it appeared to have a vast amount of information, where GIS data links appeared in many disparate sections of the site. (The GIS section was re-organized by one of the team members, George Oliver, a PhD student in Geography from Arizona State University. Later, the site was substantially edited by the CrisisCommons staff members, and now appears much better.) Finally, the skype channel seemed to provide some feel of a community gathered around a common mission, headed by Heather Blanchard from Crisis Commons. However, it quickly dawned upon the GIS group, that the skype channel was more than just about GIS, but an entire community of volunteers, all working on different missions. Furthermore, the channel was consistently joined by hundreds of people, and it was initially unclear as to who was doing what, where, and why they were participating in the channel.
After almost a full day of uncertainty, it suddenly dawned upon the GIS group that a specific mission needed to be addressed, and grouping together towards this goal would be critical to get anything done. Thus, instead of joining the skype channel, they banded to created their own Google Group, and invited Heather into the group to help keep them grounded. While they understood that this meant that Heather would have to monitor yet another group separately, it certainly helped clarify what needed to be done, and enabled them to quickly assess each others roles for each deliverable milestone. This may have been critical also in terms of forming “the team” through more direct communication through a phone conversation, rather than texts.
Over the next 5-6 days, the deliverables changed from day to day, sometimes hour to hour. Having a Google Group helped in the organization of each task by assigning different threads to each topic.
While the general mission to collect and mine relevant datasets continued on in the background, the group continued to get requests to gather new types of information to help the relief process. It should be noted that communication for these requests were coming in from a group of people at Crisis Commons who were not necessarily familiar with either GIS or Japan. In other words, any requests that had any spatial relevancy were immediately redirected to the GIS Google Group.
It quickly became apparent, however, that there was a high need for people with good command of the Japanese language, as so much was dependent on being able to search, query, and understand information streaming out of local websites. Furthermore, having contacts in Japan proved to be critical in gathering key pieces of information that would otherwise have taken days, if not weeks to acquire. On the other hand, some of the initial team members were frustrated with not being able to process data/information in Japanese and in not having a good understanding of certain Japanese standards. This was a big challenge that was underestimated from the beginning and something that would have to be taken into account for future work.
Of the six GIS volunteers, Iseki was the only native Japanese speaker, while two others (Kawano and Oliver) had passable but not fluent command of the language. Naturally, Iseki spearheaded the coordination needed to scour the web for local datasets and other critical resources, often based on information obtained through his professional connections. It soon became apparent that many Japanese GIS files would have to be downloaded from a large number of different web pages, and needed to be processed with special applications to be converted to usable formats. Iseki also contributed in what was perhaps the most critical need in this operation: educating the volunteer community on the cultural differences on basic practices around the relevant topics. Based on these needs, Elhami was able to recruit 2 additional bilingual volunteers who helped the team search for sources of valuable information.
On the other side, it became apparent that someone from our team needed to communicate directly with a group of people collecting communication infrastructure/services (CIS). Although the GIS group was constantly asked whether or not a map to identify areas in the need of portable satellite phones could be provided, the information/data coming from CIS was less than adequate due to mismatching in terms of the geographic level of information. At this point, Iseki took the initiative to join the Skype channel to work with CIS people, including Ted Han who later joined the GIS team. The product of this effort was a data file of NTT free public phones in the Tohoku region; the original file was obtained from NTT through Iseki’s personal connection, data was geocoded manually by CIS volunteers, latitude and longitude were extracted from URLs by Ted Han’s program, obtained sets of latitudes/longitudes were mapped in the Google Map by Kawano. This was a good example for a collective effort with a direct communication between those who collect data for a GIS map and those who use data for mapping.
Soon, it appeared as though the GIS group was an operation of “find and deliver”, cranking out one relevant spatial dataset after another to the international community. What had started out as a clunky operation quickly turned into a fast moving functioning group. Tasks varied from requests for additional datasets to: a need for translation scripts in python, projection data understanding and transformations, scripts to help in the task to download massive amounts of spatial data, conversations to clarify the different geographic scales and nomenclatures in Japan, and understanding and converting local GIS datasets to standards used by the international community.
Over the course of data mining, several people came on board, being recruited by Heather at Crisis Commons. All of them had substantial skills and knowledge on GIS, and two of them spoke Japanese and helped to enhance the capability to conduct data mining on Japanese Internet sites. There is no doubt that these people added substantial contributions. At the same time, the process of recruitment was chaotic and disorganized without consultation with existing members. Communication tracking continued to become a problem throughout, and even with the implementation of Google Groups, the addition of multiple team members coupled with the myriad of new requests and simultaneously on-going tasks all contributed to a convoluted, non-linear array of activities, and it soon became difficult to manage and understand critical issues such as: who initiated a task, who was the lead on each task, and more importantly, who was doing what for each task.
While it appeared that the goal was to post all our findings to the Wiki site and to the Harvard Japan Portal, these two sites were geared to provide linked information and downloadable datasets, but nothing in the way of analysis and visualizations. It was not until the group found various datasets that needed to be geocoded that a decision was made to build a portal to visualize some of the data that had been collected. By building a portal as a Google Maps Mash-up (http://gis.ats.ucla.edu/japan), it quickly became apparent that the group could add multiple layers to the site to help analyse and situate the various sources of data that had been collected.
While this Mash-up was built more than halfway along the week-long mission, this was perhaps something that should have been planned from the beginning, given the benefits of having a visualization platform that could be accessed by everyone through a single URL. Prior to having a common platform, each memeber was working on individual GIS projects on their own desktop GIS applications, sharing resulting maps as jpg’s to the group. By having a common platform on the web, this quickly allowed the team to view the information on the same platform to make informed decisions for additional tasks.
Effective communication was perhaps the single most difficult challenge throughout the volunteer mission. This is entirely understandable when you take into account that each volunteer was technically contributing 2-3 hours a day, while working from all corners of the globe. Such as it was, it was nearly impossible to set up team meetings or to coordinate skype chats with all members present. Also, the ever-changing nature of deliverables and the constant addition of new tasks and new team members made it even more difficult to coordinate. The language barrier also proved to be insurmountable, something that was adequately solved through the addition of several native speaking volunteers.
For future projects, the following are a set of recommendations put forth based on these experiences.
- It would have been better to have a briefing from Crisis Commons to let the team members have a clearer idea of goals and objectives on the mission as well as how Crisis Commons works on the project. The initial goal set based on UN OCHA was somewhat simpler than expected for the team members called for as GIS experts, and added some confusion among us, partly due to our presumptions of end products, such as maps.
- It may have been better to have a more organized strategic plan from the beginning based on better communication with Crisis Commons. Given data mining of GIS data and files from Japan as the main set goal in the beginning, the GIS team members could have set additional goals from the beginning–maybe not on the first day, but after the initial data search, taking into account the needs of refugees in Japan as well as organizations that are providing assistance and aids to them.
- It may have been more effective to assess the needs within the GIS team, communicate with a Crisis Commons contact person, and purposefully recruit more volunteers.
- Keep the open communication among the team members. In order to do so, it is important to come up with a better idea to communicate than e-mailing. A to-do task list on Google Documents is one idea as one of the team members implemented toward the end of the project. In addition, it may be a good idea to have one non-technical person who is designated to organize information.
- Establish a good bi-directional communication between Crisis Commons/its volunteers and clients and the team in order to make the team’s work efficient and effective.
- Perhaps adopting the use of project management web applications like Basecamp would be a useful addition to the workflow (Basecamp allows for communications, to-do lists, milestones, wiki’s, and file sharing; all of which were features that were at some point, requested by different members during the mission).
Despite these gaps in communication, the operation actually proved to be entirely fruitful for the team. This is most likely due to each and every member’s altruistic approach, as each member contributed their respective skillsets without hesitation to the tasks at hand. After some time, whether or not each task had immediate relevancy to the efforts on the ground became a secondary thought, and instead, a sense of community and teamwork evolved to a point where there was a mutual understanding and respect for each other’s roles.
In the end, the team had a sense that collectively, they delivered useful products towards the relief effort, and also contributed valuable expertise needed in understanding the local context. In some ways, this effort went above and beyond the requested GIS knowledge, adding a human element to the tragedy at hand, which is what any volunteer experience should be about. While some volunteers had roots in Japan, others did not, demonstrating a genuine desire to help, and ultimately find solace from such a devastating situation.