Crowdsourcing can play a powerful role during assessment of crisis. During the 2015 Kathmandu Earthquake www.kathmandulivinglabs.org (KLL) played a pivotal role in collecting data from across the affected area and making it available to decision-makers.
In this episode Nama Budhathoki, the founder and Executive Chairman of KLL, discusses the role that information and technology can play in shaping the humanitarian narrative and about the business model underpinning a small innovative organization such as KLL.
Transcript
On this week's episode of Trumanitarian, I sit down with Nama Budhathoki, the founder of Kathmandu Living Labs. KLL did outstanding work during the 2O15 earthquake in Nepal, where they crowd sourced a quick overview of the extent of the damages after the earthquake. We have discussed crowdsourcing endlessly over the past decade. It's a very appealing idea and the ability to quickly leverage a large crowd of volunteers in the affected population and collect immense amounts of data seems to be a a no brainer that it should solve many of the problems we face with assessment. However, very often, we end up with all the data being collected simply turning into a noise, rather than creating a clearer picture, because there's no analysis of the data collected and also because there often is a very weak link to the decision makers that supposedly take action based on this data. The work KLL did in 2O15 during the earthquake overcame both of these pitfalls and I think it's one of the clearest examples I've seen of successful crowdsourcing. So that's where Nama and I begin our conversation before we move on to talking about the work that they're doing reconstruction after the earthquake. And finally, we talk about what it's like to run a small, very technical organisation and trying to grow that in Nepal. I hope you enjoy the conversation.
Nama Budhathoki, welcome to Trumanitarian.
Nama Budhathoki:Thank you Lars.
Lars Peter Nissen:You are the Founder and Executive Chairman of Katmandu Living Labs, KLL. And maybe let's begin by... Tell us a bit about that. What is KLL? When and why did you find it?
Nama Budhathoki:KLL stands for Kathmandu Living Labs. We founded it in 2O13, primarily to advance OpenStreetMap in Nepal and this region.
Lars Peter Nissen:How many people are you? What's your main activities today?
Nama Budhathoki:Now we're about 20 people. Our main activities include mapping in OpenStreetMap, expanding the coverage of the map in… enhancing the quality of the map data in OpenStreetMap. In addition to mapping we also develop mobile apps and other web development, other technological solutions.
Lars Peter Nissen:And we first met towards the end of 2O15 at a lessons learned exercise following the earthquake in… that struck Katmandu that year. You found a startup in 13, you... I'm sure you start to learning some things and getting going and then suddenly this earthquake strikes. And then what do you do?
Nama Budhathoki:We are already in the collecting data, mapping and OpenStreetMap. We knew that the earthquake was coming in Nepal one day. And we also knew that the map data would be crucial, if we were hit by the earthquake. As we are doing that, we were hit by the earthquake in 2O15 as you said. You know, within the first couple of days, we realised that information was crucial, information is crucial, in any disaster. And you know what happened in the poll was.... in 2O15, was, for many rural districts, rural villages, were badly hit by the earthquake. And Nepal's rural villages, sometimes difficult to access, the road network... no road networks, there is no proper communication. So it was very important for us, for the response agencies, for the rescue workers, to understand what is happening where, especially in the countryside. You know, the people in the villages knew, of course, what they were looking for, what they needed. But the response and relief agencies, generally centred in Kathmandu and other district headquarters, did not have access to that information. So in a way, people had information, and the same information was being sought by the response agencies in district headquarters are in the big cities. So we realise, How can you collect that information, that data, from the villages and make that available to those who are already working in... you know, in helping these people. That our work is started with that.
Lars Peter Nissen:And if I can stop you there because I... one of the issues that I've had with crowdsourcing is that I think we often tend to fall so much in love with the data we can collect with these fantastic new check possibilities we have that we forget to think about the utility of that data. But what struck me exactly about your work was that it was sought after by, for example, the Army who use this information as one source into how to conduct their operations. And so how did you manage that?
Nama Budhathoki:That's a very important question. We also noticed that, you know, after the earthquake, so many different groups started collecting information. At some point, I started to feel that there is information overload, that sort of thing, you know? But we knew the collection of information is just the beginning. And, you know, the the effective use of information, effective use of data, is something... long way to go. When we started this crowdsourcing campaign, I will say, honestly speaking, we were not entirely sure at the beginning whether the information we collected would be eventually used by different agencies. But interestingly, soon after we started to collect information, the people who started to come to us and ask for that information... You know, the information on one hand was already open, because we collected the crowdsource data and made it public, through different media, through the web portal, through mobile app, you know, we even supplied a Excel sheet with a bunch of data periodically to those agencies who are interested in data.
Lars Peter Nissen:But the Army number. It's like, I'm not a military man, I don't know much about military operations, but one thing I know is, trust is not what they base their operations on. They are very careful with what they sort of take in and accept this information. How did you manage to get the Army to trust your data, the quality of your data? Did they come to you? Did you go to them? How did that happen?
Nama Budhathoki:I think that before Army started using our data, it has already created a sort of waves, you know, in the humanitarian sector, you know, there were several other, particularly youth groups, they had already studying this, using this data. So there was... there were actually two groups, you know, dedicated only to communicate and share the existence of the data we collected through a parallel source approach. So... but they did not reach out to Army. So I think Army came to know about it because there were several people... several several groups already started talking about this data. 'Hey, you know, KLL has started collecting a very useful set of data and there is a utility, you know, to leverage that data.' I met a engineer RP, Army officer. You know, I think he and his team came to my situation room, situation room means outside We were not inside the room, we were in an open space, of course. they came to us just to see, you know, what is happening, what sort of data is being collected, how useful this data might be they were interested to understand some of these basics. In the approved data, were the people involved in it. And then I had a chance to thoroughly explain this whole workflow. And I think they got it pretty quick. Then I was invited to the army headquarter. They did have information technology branch within army. The person who visited our situation room were from the information technology team, and I was invited there. And then a couple of people from the army and I went to what we call a national operation centre, that was within the Ministry of Home. We both went there and communicated, you know, and, you know, the government had a portal, and they had provided links to different sources of data. And we requested them to also include these data. So, you know, it's very interesting that it's not just me, there was there was also somebody from the army. We both went to the Ministry of Home we actually, you know, request them to put this data in the web portal, and they did that immediately. And we came back and then, you know, our conversation continued. And of course, the army did have a specific... certain specific requirements. So for example, they want the data periodically. You know, we used to publish data through the web portal as the data comes in to us from the fields, you know. But the army wanted it every three hours. So basically, we would compile this information, and we will send it to them. And then army had established a validation desk within the army headquarter. And they would validate this data before they use it. And that makes perfect sense. Because this data that has been collected from citizens, you know, no data is perfect. You know, army also... the data is one different, you know, sources. You know, they wanted to validate it with their existing knowledge. And that's how, you know, the... you know, it is starting to become a part of their use.
Lars Peter Nissen:And for me, it's an incredibly important aspect of the work you did. Because my issue with with croud sourcing often is that I it's a bit like having a 911 without having an ambulance. In other words, that there's a disconnect between very granular data being collected from a crowd of people, and in a sudden onset disaster, an expectation being raised that if I give this information to somebody, then something happens... But very often, I see crowdsourcing being disconnected from any sort of operational assets. And what you managed to do here was plug straight into the primary responder in in Nepal (the Army is obviously the biggest operational actor on the ground when something like this happens). And that made your data incredibly impactful.
Nama Budhathoki:Lars, yes. You ...absolutely. I will, you know... let me mention, is not Army. The Army was one of the major user of this data, they actually have to produce a quick report, an immense thing, how they utilise this data... You know, they also use other sources of data, of course. But there are also several other agencies will use this data. Different INGOs or other humanitarian agencies were assigned different geographic areas to respond to the crisis. And that these, these organisations were looking there specifically for the geographic area that they were tasked by the government. And our system did have a place, where people could just tell, Hey, you know, I need data from this area, from this district or from the... from this local municipality or village development committee. And the data, you know, only from that area, would go to them. And that was extremely useful for these people. So I don't remember all the organisations, I just wanted to emphasise that there are also several other organisations will heavily use this data. You know, some people even said, you know, like you said, you know, is amazing, you know, we use your data, we supply this medicine or food, whatever, you know, and it's very successful. 'Thank you so much.' If this data didn't exist, and we would not be able to do what we could do, you know. This messages from these people or organisations that actually tasked us that kept on motivating my team to continue to work for several weeks.
Lars Peter Nissen:How many reports did you collect from the crowd?
Nama Budhathoki:I think it's about... is slightly more than 2OOO reports. And what... You know, what was interesting was when the report comes in from the villages, and then we add a mechanism to update the report. For example, if people are looking for 200 tents, you know, and then that report would be there. 'Hey, in this particular, you know, community, they are looking for 200 tents.' And if another organisation supplied 100 tents, you still need 100 more tents, right? And then that information will be updated. The agency X has already supplied 100 tents, and we still need 100 more tents, you know. So for example, some of the agencies, you know, probably they don't have, you know, capacity to supply all the need. You know, everyone can supply partial, you know, need. So yeah, again, I think I don't remember the exact number right now, but this is is anything between 2OOO and 3 target?
Lars Peter Nissen:And how long did you continue this work of collecting the immediate needs right after the woe? How long was that phase of your work?
Nama Budhathoki:It went for a couple of months.
Lars Peter Nissen:So, in the initial months, you collect around 3OOO reports from villages about the immediate needs created by the earthquake and then what do you do once focus starts shifting towards reconstruction? How did you deal with the reconstruction phase?
Nama Budhathoki:We work on the response for the first... five to six months. You know, of course, during the first couple of months, it was... we were very active. And then as time passed, you know, it slowed down, we were also tire and exhausted. And then our focus shifted to reconstruction about six months after the earthquake. Our major contribution in reconstruction was that we helped the government to collect the massive damage data. Our government needed the assessment of the use of the buildings in the villages in districts, you know, in order to plan their reconstruction. So they decided to use technology in order to assess the buildings and collect data from the field. So I'm talking about a million buildings, by the way. So that's a big number. They made a very interesting decision and I really respect our government for taking that decision. Traditionally, they used pen and paper to go to the villages into something like census. Collect data in paper, bring the data to district headquarters or Kathmandu, and convert that paper data into digital form. You know, it would very easily take take several months, years, you know, but there was no time. You know, the thing is, unless the data is available to the decision makers, they cannot start planning for the construction. So, they wanted to use technology. So idea was, they would recruit about 3OOO engineers, because you need engineers to assess buildings, right. And then each of these engineers will be given a tablet device, loaded with a software, and they will go to the villages, visually inspect the buildings, interview the owner of the building for about 30 to 40 minutes and record all that information in a tablet. And they would immediately upload that information and that... the data would come to Kathmandu, you know, in the government server. So, that is then becomes available to decision makers from next day, that was an idea. But it was at the beginning you know, a lot of people did not believe that that will happen. Because in villages in Nepal, there is no electricity. How would you charge your device? There is no internet. How would you send your data? And then we did a lot of discussions. But anyway, I... you know, Nepal took that decision and we had an opportunity to be the part of this. You know, KLL was selected to help the government and all the technology related, you know, activities. So designing app, joining engineers to use the technology, was the send the data to the server, make sure data sits currently on the server, you know. And then help the government officials in data validation, data processing, you know... So we work with, you know, different organisations. The most of them were from the government and this is very interesting... You know, Lars, I must tell this. In about four months, we collected data for more then, I think, 8OOOOO buildings, from 11, you know, badly hit districts. You know, for months internet housing buildings, about 3OOO engineers, you know, is unbelievable.
Lars Peter Nissen:Yeah, I just did the math, it's some somewhere around six and a half thousand buildings per day over that period, on average. That's truly impressive.
Nama Budhathoki:Yes. And is considered to be one of the biggest mobilised survey in the world in those days, you know. Biggest in the sense that everyday you did have 3OOO engineers in the field, at any moment, there is coming to the server continuously, and it is not only a text data, we also collected pictures. You know, we had to collect 10 pictures per building. You know, when the engineers assess the building, they would assign a grade.Grade five means completely damaged, one means nothing has happened to the building. So in order to substantiate, in order to produce evidence against that, you know, grade, you know, they also had to provide... take a picture and upload that picture. And the pictures take more bandwidth than takes data, and we had a massive salaries to bring those pictures. And we did a pretty interesting few workarounds, you know, for the developing country, you know. If... when people think... you are in Geneva... if you are thinking of this project from Geneva, you wouldn't probably worry too much about the internet bandwidth. You would operate with that conception, and it would very easily tell, you know. But we knew the path, we need the possibilities pretty well. And we had started thinking about that from day one. And, you know, that worked.
Lars Peter Nissen:It is a truly remarkable story. And I think it's worthwhile just going back and remembering that you founded it KLL in 2O13, so you're less than two years old, or about two years old, when this happens. And and it is... a lot of credit to you and the team for actually pulling this off. I'm sure that in Nepal, this work has been greatly appreciated by the government and your clients. And so, what's next for yo now? You you come out of the earthquake some years ago with with this very positive experience and what... How have you developed KLL since then?
Nama Budhathoki:I should tell you one thing, frankly. You know, we haven't been able to put a consented, you know, effort to deeply reflect, learn, and prepare for the future disaster in Nepal and other parts of the world. You know, there are people from the academia, very interested in these work. People are reading journal articles, you know, and there are several other discussions. But, you know, I had expected that this thing would inform, you know, our preparation for the future in a substantial way given the fact that a lot of people liked it. As you said, this does work, both the crowdsourcing work we did, you know, during the response phase, and also this damage assessment work that we supported the government in the reconstruction, I think a lot of people have appreciated this. And they... you know, some people find it really, really, you know, impressive work. But I think it's a sort of a missed opportunity. It's not that, you know, the people haven't... you know... I don't know how to put it. You know, we could have built of it. We could have developed. You know, and what we did in 2O15 Nepal earthquake is not something perfect, but it gives us a pretty good... you know, a groundwork that to discuss, to debate, to build. You know, that requires some effort. You know, KLL has been doing that. You know, I think we have inspired other organisations, both within the Nepal and view in Nepal, I'm pretty happy with that. That's, one side of the story. And now coming back to your question, what KLL has been doing, you know, KLL is not limited, just in disasters. You know, our work is, you know, sort of cross-cutting. We have also contributed some of the governance projects. We also do some research with the university. You know, we also continue to do OpenStreetMapping, which is our flagship programme, you know. So, you know, we... you know, during the last a couple of years, we did open... you know, the OpenMapping in Nepal, in India. You know, is a slightly say humanitarian work, there, again, the preparation for disaster. And once you have a data, you can also use the data beyond disaster, you know, once you have impressive map of a city, you can use the map when there is a disaster, and you can also use that map for other planning activities.
Lars Peter Nissen:But it sounds to me like you feel that there is a missed opportunity. Why do you think that is? Why did people not queue up to invest in this sort of preparedness, which obviously was very impactful in 2O15?
Nama Budhathoki:I think the work we did in 2O15 earthquake is not, you know, quite there, you know, at the Ministry of Water.. I will say a traditional disaster response. And as you use the word, you know, the humanitarian model is a little different, right? And then... so, you need to stage, you need to go beyond your comfort zone, you need to try to understand what happened, you know? In many cases, people want to continue what they have been doing in the past. You know, that's relatively easy, that's comfortable. You know, when things are little hard, you know, that might be interesting. And people say... and I talk to people, pupils, here, that's pretty interesting. I amazing, you know? That's one thing. And, you know, being able to, you know, to understand that and integrate in your work, requires a different level of courage, a different level of determination. You know, that's, I think, what is missing.
Lars Peter Nissen:So basically, the mainstream actors have not managed to engage with you in a way that has led to really a sustained effort to create a higher level of preparedness,
Nama Budhathoki:I would agree with you. And it's not just, you know, Kathmandu Living Labs work is any new innovations and new ideas. You know, people are... majority of the people are reluctant. You know, they don't want to take risk. You know, they don't want to put a sufficient effort to understand, to internalise, these new things, you know, partly because they might be busy into, you know, doing what they have been doing. You know, I really don't know. You know, that's a different topic of discussion. Why don't people... Why do people continue to do what they do? You know, in... that's what I see in majority of the cases, but there are people, you know, they're looking for new ideas. But you need resources, you know. You know, do these people have resources to do that? Do they have that luxury? So there's not one question, Lars. There are a bunch of these questions around this conversation.
Lars Peter Nissen:I think we need to revisit the humanitarian business model, if you want. If I can speak a bit in economic terms. I think, on one side, clearly what KLL did during the earthquake creates much value for the community as such, but at the same time, I can see a lot of organisations in similar positions as yourself (sort of small, technical, very capable service providers) struggling to break into getting money out of the mainstream organisations. I actually think that there is sort of a market failure there, and that we need to revisit that. Because if you are capable of delivering the level of services and the quality of information that you deliver during the earthquake, then it pretty much should be a no brainer to also throw some money at that. But we can see in your case, but I think in many cases, the difficulty for the mainstream humanitarian organisations to let go of part of the value chain. There's a strong tendency, I think, to want to do everything as as a response organisation.
Nama Budhathoki:Yeah, you're absolutely right. You know, it could take some time, Lars. You know, the conversation has been started. You know, we need to continue to, you know, continue this conversation. We need to continue to educate people. We need to continue to demonstrate the the value of doing things, you know, differently. We have to, you know... we have to... we have to debate that the existing model probably is not working to the level that we all would like to see, you know? We're living in a very different context, in a technological context, things are changing, you know, in many fields the rapid development. And, you know, we just can't already know that, you know, that will not be a good thing for humanity, I guess. But I know, you know, it takes time. You know, the change is inevitable, is going to be there. You know, when also depends upon, you know, intensity of our conversation, our work, you know, our enthusiasm, you know, to put people in the seat. And then I'm very optimistic. You know, and there are actually examples around the world, especially indications that there will be change, but, you know... but I wish, you know, it happens sooner than later.
Lars Peter Nissen:Nama, thank you so much for coming on the podcast and sharing your experience. You're doing fantastic work. I'm so impressed with KLL and the progress you've made over the past years. I wish you all the best luck for the future, and hope to have a chance to engage with you in person one day again.
Nama Budhathoki:Thank you, Lars. I thoroughly enjoyed this conversation. My last quick note is, you know, over the last, you know, couple of years, you know, after the earthquakes, we've been continuing mapping in OpenStreetMap. It's very likely that I will be doing some mapping work in Nepal, and few other countries in Asia, you know, to map the secondary impacts of COVID-19.
Lars Peter Nissen:Nama, I look forward to seeing the outcome of that. It's an incredibly important piece of work. And I think it's what we all struggling with these days, is understanding what are actually the secondary impact of the pandemic? How can we get evidence to actually tell the right story about the way in which humanitarian and development outcomes are being reshaped by this crisis.