This is the third and final post in a 3-part series discussing a partnership I initiated between Tulane University, the American Red Cross, and OpenStreetMaps, in which students in my undergraduate IDEV4100: ICT4D course participated in real-world mapping for OpenStreetMaps using satellite imagery, to support development and disaster preparedness initiatives being conducted by the American Red Cross in coordination with local host Red Cross and Red Crescent National Societies.

Outcomes Beyond the Base Maps

For Robert Banick, GIS analyst at the American Red Cross, through our emails and conversations, I know that at times fitting our class sessions into his schedule could be challenging, though we tried to be as flexible as possible, his travel schedule was often demanding. This also meant that his ability to Skype with our class could be challenging, particularly when he was in the field with one of his partner countries, such as when he was working in Chile or Nepal, and we were faced with spotty Internet access and connectivity issues on his end, which compounded some of the troubleshooting issues we were having with software, when one of our technical advisors became less accessible… though it did make for some teachable moments! Another complication was that I could never be certain too far in advance what the exact number of students enrolled would be, which made it difficult to tell how many students we could have working on a particular map, in turn complicating his decision-making somewhat. In addition, I couldn’t guarantee an extended amount of time for the students to spend on the project, because of the other scheduled assignments and to ensure respect for student time, I tried to make sure they knew what kind of time commitment was expected from them in advance. Robert was therefore managing a formula of X number of students times Y amount of time = Z volume of work, while fitting into one of the current needs of his partner Red Cross or Red Crescent National Societies.

Screen Shot 2016-03-16 at 5.12.56 PM

Figure I: Close-Up View of North Jakarta in OSM, February 2015

It is most evident that the students’ work continues to be built upon in Indonesia, when you look at the state of the maps in Jakarta and Depok now, nearly two years later. The maps there have continued to be updated by local volunteers in subsequent stages of the project, with street names and the addition of symbols indicating types of buildings such as mosques and hospitals. (see Figure I) Based on my conversations with him since then, most of the maps our students generated have been used, though unfortunately some have yet to be utilized (I suspect those that haven’t been utilized yet were those in Nepal, based on the limited progress the students made on those maps before they were pulled onto the Haiti project). But according to Robert, “that’s to be expected with a novel pilot approach like this.”

Our experiences with the IDEV4100 students significantly impacted the development of the Visual Tracing Guides that Red Cross has been using for their subsequent crowdmapping efforts. They have put out public calls for mappers to help with emergencies such as Typhoon Haiyan in the Philippines (see below) and the Ebola outbreak in Africa, and the tracing guides they provide to volunteers for these efforts is clearly an outgrowth of the ones developed for our students. They are also relying more heavily on iD Editor as the go-to software platform in areas where the Internet connections can handle the traffic. JOSM is still a more reliable platform in areas where there are less reliable internet connections, because it is easier to download a file, spend time offline doing all of the editing, and do one single upload of a file with JOSM rather than the dynamic, constant edits that you would do in the web-based iD editor, but the simpler interface of iD editor makes it more appealing for short-term mappers with less experience who want to get started quickly and have reliable Internet access and high bandwidth.

Robert’s travel schedule took him to Indonesia during the same semester our mapping project focused on Jakarta and Depok, in May 2013, where he was working with the Indonesian Red Cross and HOTOSM on the next phase of the mapping project. Our work (along with the work done by some students at George Washington University – see Box #1) was used for American Red Cross-funded institutionalized mapping projects to assist with disaster preparedness. The Indonesian delegate he spoke with and showed our maps to was very excited about them, and we received some excellent feedback that informed how we prioritized some of the features we mapped.

That fall, when Typhoon Haiyan hit the Philippines, the Red Cross led OSM efforts to coordinate the use of OpenStreetMaps and volunteer efforts around its use to support relief and response activities in the wake of the disaster. To date 1,679 individuals have volunteered work on the Haiyan OSM project, contributing nearly 4.8 million map changes. Robert himself spent time in Tacloban following the disaster, working with the team creating maps relating to housing and sheltering. When asked how he knew that the maps were making a difference, he responded: When I was in Tacloban, I ran into a Red Cross team handing out relief supplies. They told me that OpenStreetMap—which we loaded onto their GPS devices as they deployed—was super useful. The maps saved them from getting lost or wasting time when they had to reroute off damaged roads. They were able to give directions to Filipino drivers. It all leads to more efficient delivery of supplies to people affected by Typhoon Haiyan. The effort and attention that is paid to collecting data for maps is super important. Desk jobs sometimes get short shrift, but the output we produce makes a huge difference.”

Since working with us, the American Red Cross International Services GIS team is significantly scaling up their crowdmapping efforts, in partnership with Médecins Sans Frontières‎ of the United Kingdom, the British Red Cross and the Humanitarian OpenStreetMap Team under the banner of the “Missing Maps” project. According to Robert, “Our successes, failures and learning from working with your classes have significantly informed the setup of the project.”

Outcomes for Tulane’s International Development Students

Screen Shot 2016-03-16 at 5.14.05 PM

Figure J: Tulane Student Involvement with Red Cross & OSM

The reaction I heard most often from the students when they heard about this activity was how much they appreciated being able to work on a project that was taking place in the real world. The opportunity to have an assignment based on a real-world project, something affecting people in a tangible way instead of a simulation or an exercise, something outside of the university setting, was invaluable for them.

“While this project is and was very daunting to me – being that I am untrained and very under-qualified to be working on something so important – it at least gave me a chance to put something I have read about into practice. That is not something that a lot of students get to do. I find it particularly important in the field of International Development, where it is a common misconception that IDEV experience can only be gained through something like volunteering abroad in a place like Sub-Saharan Africa, that those studying and those interested in the sector gain hands-on experience working on something in it.” (Student 144)

During the mapping assignment, however, most of them experienced a great deal of frustration. As part of their assignment, they all submitted final reflections where they were encouraged to discuss the challenges they personally encountered when working with the software and the OSM Tasking Manager, and throughout the activity. The four most common challenges mentioned were:

Software Difficulties

Learning a new software package can be challenging, and how easily it is learned varies greatly from one individual to another. The student encountered issues with downloading the JOSM software, installation issues, uploading errors, problems learning the “ins & outs” of how the program itself worked, and other challenges. The interesting thing about student reactions to the software, whether it was to JOSM or iD editor, was that their reactions did not necessarily relate to their self-identified comfort level with technology, as in whether they saw themselves as “tech-savvy” or not. Some students who considered themselves tech-savvy had “major” issues with the software, while others who claimed they were tech averse or “not good with technology” caught on quickly and became comfortable with the software quite easily. From an instructor perspective, and with the benefit of retrospect, it seemed to be more related to a student’s attitude toward the project, in that some students came into the project, or to the class more generally, with a more open mind, more eager to learn about the topic, and those students, regardless of their level of tech-savviness, are the ones who performed better with the software overall, learning from mistakes and taking it one step at a time, asking questions and understanding it as a process rather than becoming overly frustrated. Others became quite immediately frustrated with the software, and were more likely to conclude that the software was “flawed”; an interesting conclusion after using it for such a short amount of time.

                 “I can say that I tend to get lost easily when it comes to technology and software programs besides the stuff that I use everyday but I thought that by halfway through tagging my first square I really had a grasp of the programming involved with the project.” (Student 101)

“Since we did not receive a very thorough tutorial on using the program prior to starting our grid squares, we were not aware of all the functions available to us. I learned as I kept working, but it took a lot of messing up and working in inefficient ways to get there. It was certainly rewarding to trace my progress and observe how I could learn to master using a program that at first made no sense to me.” (Student 145)

“Low Quality” of Satellite Imagery

Although the satellite imagery we were given access to was the best available in the area at the time, in many places it was still too low-resolution to be able to tell in great detail what we were seeing. While in most cases it was good enough to trace the major formations and create base maps, it was difficult to tell where one building ended and the next began, in some cases where the buildings ended and their shadows began, what areas may have been fields and what may have been structures, and in countries and areas the students had never been and weren’t familiar with, it was difficult to distinguish one thing from another. Being used to Google Maps and zooming in so far that you can see your neighbor’s face sitting on their front porch across the street in New Orleans, this was frustrating for them. Many expressed strong reservations about tracing and labeling spaces and structures when they were unsure about what they might be.

Lack of Familiarity with the Area

By and large, the students were mapping countries and cities they had never visited, which compounded the issue of low-resolution imagery discussed above. Without understanding what “normal” houses or businesses looked like there, it was difficult for them to make judgment calls about how to label particular structures. They were forced to tag structures and landforms with the most general tags available to them “building: yes” instead of “building: commercial” or “building: residential” for example, which to them was frustrating because they felt their work was incomplete. It was also, on some level, boring because all of their buildings were being tagged as the same even when there were clear differences between them.

Real World Pressures

While nearly all of them cited working on a “real world” project as a positive aspect of the assignment, for a number of them this seemed to add an extra dimension of stress to their work, in the sense that they felt an additional pressure to have their work be absolutely correct despite the fact that they knew little to nothing about the geographic areas they were mapping (and were not expected to). We explained to them on several occasions that their mapping efforts were simply the first of several stages of mapping that were going to be conducted, that they were truly creating base maps, outlines of the general land uses, roads, buildings, and major elements that would then be followed up with on-the-ground, local surveys, conducted by local partners. Their work was not designed to be the final draft, and lives were not hanging in the balance. Some of them, however, remained quite concerned and the frustration carried over into their feelings about the entire assignment.

“The neatest part about this project was clicking on my username to see the change sets that I had made. What had previously been a blank square was now part of a community with a river, a main road, and many houses and huts. I feel proud of my contribution because I know that these maps are going to facilitate service delivery and planning in Uganda. Nonetheless I am concerned that my inexperience in applications like this resulted in errors. I worry about the accuracy because I know that these maps are actually going to be used by an established organization that has the ability to help so many. I do not want to be the cause of inaccurate delivery routes or poor planning.” (Student 104)

One of the most significant lessons the students could learn from this process was to better understand from a personal perspective, this process of learning a new technology and what implications that can have for International Development generally, and for utilization of Information & Communication Technologies (ICTs) within IDEV. Some of them drew lines between their personal experiences and what it would be like for individuals in developing countries to learn new technologies without the benefit of growing up in the kind of technology-heavy environment that our students might have in the US.

“If I did not use computers on a daily basis and had not been educated in how to use them, I would have most likely had difficulty using the program. This is due to the fact that it is prone to automate certain tasks such as binding nodes together which if not corrected have to be fixed before uploading. The fact that utilizing the program requires a moderate to high level of expertise in technology poses a challenge because it makes it difficult to use the knowledge of the locals of developing areas hat need to be mapped through crowdsourcing.” (Student 151)

The process of this assignment also enabled us to have long conversations as a class about the importance of sufficient training and support for technology programs, the importance of local knowledge and local buy-in and participation, a broad and lively discussion of the value of maps in everyday life and for various sectors of development (such as health or emergency preparedness, public safety, governance, or environmentalism), and public/private partnerships and the involvement of different stakeholders. This was where the most value of the assignment lay, from my perspective, aside from the value of students having a truly hands-on experience with technology. To be able to take that experience and have the students extrapolate from it how the design and implementation of ICT4D programs might better occur in development, that was the most valuable conversation we had each semester.

Conclusions & Recommendations

Over the course of three semesters, the students enrolled in Tulane University’s IDEV4100: Information and Communication Technology for Development (ICT4D) course participated in a real-world, hands-on crowdmapping project in partnership with the American Red Cross, and in coordination with Humanitarian OpenStreetMaps, to create useable base maps in program areas of partner Red Cross Red Crescent National Societies in the developing countries of Uganda, Indonesia, Nepal, and Haiti. The purpose of this mapping project was to provide useable base maps which would serve as a first step in a series of mapping efforts in each area to provide more reliable information for first responders to utilize in the event of floods, fires, or other natural disasters in these target areas.

The engagement of these students as quasi-volunteer crowdmappers, linked their efforts to the existing “digital social innovation” partner chain. In order for these efforts to be deemed successful, they needed to meet objectives on both ends. For the students and the course, the activity would need to meet learning objectives and provide an opportunity for students to engage with concepts they encountered in their classroom. For the Red Cross partners, it needed to at least have the potential to generate a volume and quality of work that would be beneficial to them as they entered the next phase of mapping. Based on these general criteria, the efforts can certainly be deemed a success. Students were given the opportunity to learn a new software platform and engage personally with new technologies, learning valuable lessons about the processes of training and learning, as well as software design, satellite imagery, mapping, crowdsourcing, the importance of local knowledge, and other significant concepts in ICT4D and international development. For the Red Cross partners, the volume of work generated, particularly after the first semester, proved to be worth the time and effort involved in the partnership. Indeed, the lessons learned and the experience of working with this type of group, including the development of the Visual Tracing Guides and engaging short-term volunteers with little to no experience, is part of what enabled the Red Cross to successfully engage volunteer crowdmappers for emergency efforts to map after Typhoon Haiyan in the Philippines, for the Ebola outbreak in Africa, and for their new Missing Maps initiative.

Unfortunately, the model was partially dependent on a few elements that were unsustainable. As an adjunct instructor, my term teaching the course was temporary, and once I was no longer the instructor, the syllabus changed to reflect the expertise of the new instructor, and no longer included the old syllabus I had created, as is the norm with the design of this course, it changes over time based on the rotation of the instructors and their differences in technological focus. However, the model is replicable, and once the model is defined, it is somewhat easy to duplicate in another setting, even with an instructor who doesn’t have much training, as long as it can be made relevant to the course content.

Despite these flaws, it seems clear that this type of partnership and integration of a group of university students enrolled in the ICT4D course into the existing partnership of Red Cross and OSM is something that would not have been possible without the digital technologies available today. These digital tools are enabling social engagement and partnerships in new ways, with the end results focused on social programs and development results.