After the March 2011 earthquake and the explosions at the Fukushima nuclear power plant, TEPCO and the Japanese government tried to assure us that everything was just fine. The repeatedly insisted that there was no serious danger posed by the radiation.
Not very many people believed them.
Reliable data from fallout areas was sparse at best, and many Japan residents doubted that the government was telling the truth in the first place.
It was in that environment that Pieter Franken and his team created Safecast. Safecast began as a small group in Japan with home-made Geiger counters making their reading available to everyone. They have now grown into an international movement involving private citizens, universities, non-profit organizations and government agencies.
Pieter also explains why environmental science will look very different ten years from now.
It’s a fascinating discussion, and I think you’ll enjoy it.
Podcast: Play in new window | Download | Embed
Subscribe: Google Podcasts | Stitcher | RSS | More
Show Notes for Startups
- Why Japan’s disaster preparation failed
- Why you need high-resolution and high-density radiation monitoring
- Why citizens do not, and perhaps should not, trust their governments
- The advantages of creating a DYI kit rather than a product
- How to maintain data integrity for crowdsourced efforts
- Why both pro-nuke and anti-nuke activists opposed Safecast
- How governments have reacted to alternative data sources
- Safecast’s plan to win over the scientific community
- The future of citizen science
Links from the Founder
- Everything you wanted to know about Safecast
- Connect with Pieter on LinkedIn
- Follow him on Twitter @noktonlux
Transcript from Japan
Disrupting Japan, episode 89.
Welcome to Disrupting Japan. Straight talk from Japan’s most successful entrepreneurs. I’m Tim Romero and thanks for joining me.
You know, crowdfunding and crowdsourcing in Japan largely gained in its popularity in projects related to the massive March 2011 earthquake, and ensuing tsunami, and the release of radiation at the Fukushima Nuclear Power Plant. In fact, longtime listeners have heard the founders of some of Japan’s largest crowdfunding and crowdsourcing companies explain that breaking away from this image of crowdfunding as a social good was something that they had to overcome before their startups became truly successful.
Well, today we’re going to sit down with Pieter Franken of Safecast, one of the earliest examples of widespread crowdsourcing in Japan. And we talk about how they’ve grown from a Japanese patchwork solution to the leader of a global movement. After the Fukushima nuclear disaster, people throughout Japan were worried about radiation. TEPCO, who operated the facilities and the Japanese government assured everyone that things were under control and that everyone was perfectly safe. As you might imagine, however, most people were highly skeptical of these claims. The radiation data just wasn’t there or it wasn’t being shared with the public or it wasn’t believed when it was shared with the public.
Pieter and his team started Safecast to make sure that lack of information and lack of transparency would never happen again and they began building low-cost Geiger counters that people around the country and then around the world could use to measure their local area and then have all that data uploaded into the cloud and made available for anyone. It’s an amazing story and it’s one that Pieter tells much better than I do. So let’s hear from our sponsors and get right to the interview.
Tim: So I’m sitting here with Pieter Franken of Safecast. You guys make an open environmental data collection system for everyone but I think you can explain much better than I can what it is.
Pieter: To explain what we’re doing, best thing is just to go back in time a little bit. Exactly six years ago in March 2011, we all witnessed the big earthquake and as part of that, the Fukushima Daiichi disaster. And it’s really where Safecast started. It started from the need to know what was happening in terms of radiation exposure in Japan. Also, the folks that were here will remember very well there was lots of a concern about that but also the biggest problem was there was no data available. Whatever the authorities said was very sparse and most of it was not useful, nothing was really from Fukushima itself. We tried to solve that problem.
Tim: Also as I recall that there was a lot of people that just simply didn’t believe what the government was telling them.
Pieter: Yeah. Lots of people didn’t trust TEPCO’s own assessment. But the real problem I think in retrospect was there was very little information available. Whatever was available was not disclosed. By the sheer fact that information was not available, that created even a bigger distrust. That was really the space where we decided to start Safecast. We said, “We must know what is happening in the environment.” Radiation at that time was the critical driver. We said, “How can we make this as open and transparent as possible?”
We did all kind of things in the beginning. We didn’t start with the soldering iron and building things. We actually thought that information would be available yet it was just hard to find or not easy to digest. So we started to look around on the internet hoping to find all that information and then put it on a website and share that; however, that didn’t really fan out. We didn’t find information. Most of it was from some universities that all were in Tokyo or measuring something that wasn’t relevant. Our MEP was very empty after looking around for two weeks, so started to rethink.
Tim: Did the Japanese government have a unified national data collection system for radiation?
Pieter: Yes and no. They actually had a system. The system was called ‘Speedy.’ That system was originally designed to predict where the plume would go after a nuclear disaster. So it had sensors and it had a huge mainframe with a huge piece of software on it. Then we’d read all the sensor data and then churn out, project the trajectory of the plume.
The key thing is that system was there. It did operate but it is not a publicly accessible system. It’s not that you’re going online and see what is happening here. So the information was exclusively available to Japanese government only. And for all kind of reasons, they decided to ignore that information. So this is one of the big parts of the whole accident that went horribly wrong. So they didn’t look at that information and they started to improvise from that point onwards.
To answer your question, yes, there was a system not publicly available and effectively was not used. Today, that system has been decommissioned. Today, Japan doesn’t have that system anymore. So we actually have less than what we had before the accident.
Tim: Well, hopefully with Safecast we’ve got a more thorough data than we had before.
Pieter: Right. So we realized that information wasn’t available. We couldn’t put on a website so we said, “Okay. Maybe we can crowdsource the data.” The initial idea was to buy Geiger counters and give them to folks so that they could collect the data and tell us so that we could publish that. We did try to do that but shortly after the accident, all Geiger counters worldwide sold out because everybody in Japan needed one. But what we didn’t realize was that they would be sold out for 6 to 12 months because the supply chain for this is not – they’re not iPhones or something where millions can be produced on demand.
Tim: Yeah. I’m sure there’s been a slow steady demand for them.
Pieter: Yeah, yeah. Exactly. So the factories that are churning them out were not going to scale up and were not very easy to scale up because the actual sensors are not so easy to manufacture. So the net-net was there’s no Geiger counters. We had actually run a kickstarted campaign by that time. So we got some money and then we started to rethink. We don’t have Geiger counters. What do we do next? So we had a few Geiger counters that we had. By that time we had maybe 10 or so. The idea was born that maybe we can do what Google does, kind of a street map view of radiation by driving around. That kind of started into something that we thought we could execute. We did a trial. We had a truck driver go up, David Kell. He was one of the volunteers going out by truck. We gave him a Geiger counter and an iPhone and we said, “Whenever you stop, just take a measurement and after you have taken a measurement just upload it.” As he was travelling up north, he kept on sending these pictures up and we could see what was happening. So it was kind of the real first version but it was just to see that you know, would this work? The second version of this was a group of Keio University students, and we were more structured. They were driving up and we said, “Every 5 minutes, take a picture of the Geiger counter.” They kind of duct taped the Geiger counter onto the car on the outside and they were sitting inside. So they could see the screen of the Geiger counter and every 5 minutes they snapped a picture.
Tim: Yeah. In principle, that’s still kind of what you’re doing.
Pieter: That’s what we’re doing. The only thing we did is we automated the process. We had a bunch of people that were at Tokyo Hackerspace. We all got together and we started to say, “How can we make this thing with Arduinos and make it more automated?” So we put the whole thing together. We chose a box to put it all in which looked kind of a bento box. We hooked up a computer and that was kind of the real kind of the first thing we got going.
Tim: So you basically designed a do-it-yourself Geiger counter kit.
Pieter: Right. Well, the kit came much later. We took a real commercial Geiger counter because we didn’t have anything else. So we took that, we interfaced it with an Arduino, hooked it up to a PC. And on the PC, we wrote o software that every five seconds would read out the value from the Geiger counter and store it on the hard disk. So it was really the first thing we did. With that, we went to Fukushima and did our first measurements.
Tim: Okay. How long after the earthquake was this?
Pieter: So this was about six weeks after the earthquake, roughly. We knew that time was working against us because as the disaster was progressing, there was no information. But bringing information a year after the accident is not really interesting, we wanted to know now. And more specifically, I have a family here. I have a young daughter. I wanted to know of what my risk was and should I move away from Tokyo or what is happening here?
Tim: Six weeks after the accident, there were still a tremendous amount of concern.
Pieter: Absolutely. Even today, the concern still continues for specific areas, so we really wanted to move as quickly as possible so that people would get the data as early as possible. The first so called ‘B-Gaigi’ which stands for Bento Geiger Counter System on the road was towards the end of April 2011. At that time we drove up to Koriyama City and we started measuring there. We collected the data. We also immediately were on the ground, we met people. We started to see that the levels were way higher than what we had thought they would be based on what NHK News and other news sources were telling us. And then we started to publish our first radiation map.
Tim: At this point, you just had one unit?
Pieter: Yeah. At this point, very quickly after we built the first versions of this, we eliminated the computer, the laptop, and we used an Arduino self-contained system. We made it waterproof and we also made it super easy to use, basic on and off switch and it would just record or not record, so no need for people to set computers and stuff like that. We could hand them out to volunteers in Fukushima. That was kind of the next major way, and it was still kind of clunky and it still required a commercial Geiger counter to be in the box. So they were relatively expensive and we had shortage of supply, et cetera, et cetera. But we had, I think, tens of volunteers over the first three months that started to measure. We kept on churning out these boxes by putting it together here in Tokyo and we started to get more and more data. I think in the first three months we had about half million data points collected. And we started to really kind of see what was happening in Fukushima. The contours or where things were high, where things were low. We really started to realize that it was very blotchy, it wasn’t really predictable based on distance. And even in cities that were exposed; inside the city you would have higher areas, lower areas. So it really started to become clear that in order to really get an idea what was happening, you had to go measure everything.
Tim: So you say a higher area and a lower area. You mean like different parts of town or different corners on the same street?
Pieter: There were corners on the same street could be quite different. What we found out is that it had a lot to do with characteristics like a big roof could a concentrator or fallout into one little corner that could create a much higher hotspot. Also, the soil or the road or the material it fell on had an effect on how high you would measure. So we literally could walk around short area and have very different measurements.
Tim: Wow. Okay. So you really do need very high resolution, you need a tremendous amount of data to get anything meaningful out of this.
Pieter: Yes. Yes, absolutely. And you know what, it’s really important to understand this, is that the people are still living there. For them this was really, you know, real reality. Is my street, okay or e not okay? What should I do? Whatever. We started to really work with people in the communities that really wanted us to help them out to measure their street, their workplace, their school, their hospital. Then the secondary thing is that people that after all were living in affected areas. Tokyo was affected but not so much effect as Fukushima. It didn’t really matter where you were, people were equally concerned. The reason for that was we just didn’t know.
Tim: Right. Well, I mean people didn’t trust the data they were being told. They didn’t understand what the data meant anyway.
Pieter: Right, right. But also the same thing in Tokyo, whatever the data that was available was – few locations only in Tokyo were measured and that was official measurement. But we knew in Tokyo there were hotspots as well and inm Chiba there was a hotspot and stuff like that. It became equally important to measure Tokyo or Kyushu as it was to do Fukushima because the concern was across the country. So 2011 summer, we kind of made the commitment. Our goal was to measure everything; that solves the problem. Japanese government at that time had their hands full measuring inside the exclusion zone and wasn’t measuring outside. So it was a great complementary thing to do.
Tim: So up until this point, the people who had measuring? Had it been affiliates or – it wasn’t a true crowdsourcing community yet, right? Or was it just turning that corner?
Pieter: Well, what is true crowdsourcing is we didn’t direct people to go do anything. People would raise their hands and we would provide them equipment to go measure with. From a Safecast point of view, at that time, we were maybe maximum 10 people in Tokyo that were working on it. We had way more people in the field doing all the measurement.
Tim: So when did you develop the first Geiger counter kit?
Pieter: Yes. The kit was kind of born about a year after that. By the time we already had maybe four or five iterations of the system that we had built initially. By the time it had become quite sophisticated. But one of the big things was some people wanted to own it and we were only lending it out. So the other thing was just to form a factory. If we could make it smaller, wouldn’t it be great. If we would be able to avoid the Geiger counter in there, we could bring down the cost. So we experimented with that and we found a form that we liked and then we iterated a little bit further and then the idea was born. If we can make this a kit, then people could build it themselves. So that would take away the problem of manufacturing which has a huge lead time.
Tim: But didn’t you run to the same problem before of the supply chain that makes the detection units?
Pieter: Yes. So this is about a year after the accident. Around that time, the supply chain has sprung into gear and also we were lucky that one of our advisors was running a company making those and he could help us source the equipment. The first batches of these kits were built in workshops. We basically got 10 or many people in the same room and we would teach people how to build them. Then over time when we had a company that made it into a kit that you could buy in Amazon. So suddenly outside of Japan, people started to buy them. Fast forwarding six years later, we have many of these in use worldwide now in many, many countries where people are using these.
Tim: About how many units are there worldwide?
Pieter: My guestimate is around over a thousand growing rapidly that’s why it’s my guestimate. These are now being used in many, many countries. I think we have counted way over 90 countries that we have data from.
Tim: I imagine most of the coverage is in Japan.
Pieter: Not anymore. About three years ago, we got lots of volunteers in the US that started to measure a lot. After that, Europe came online big time and we have lots of volunteers in Europe now that are measuring fairly seriously. Roughly maybe still over 50% is Japan but the rest, over 40% is from other countries now.
Tim: Okay. So the device itself, it’s fairly expensive. It’s $600 or so for the kit. It’s not a trivial matter to decide to do this.
Pieter: Yes.
Tim: Who are the people and organizations that decide to participate? Are they universities? Are they just –
Pieter: There’s no simple answer. That’s very varied. We have, of course, individuals like ourselves but we have companies that for some reason are measuring other things already on the street and they find it interesting to participate and contribute but also learn a bit about open hardware and how does this whole new thing work? We have universities that are participating. We have local cities in Fukushima that are participating. We have a non-government organization that are participating in the project. We also know that people buy them as a group activity, so people chip in. Everybody 50 bucks and they build one together and they share it in the community.
Tim: Okay. Let’s talk about building a community and the value of the community. Data is great but data by itself doesn’t have as much. So why is this good? What problem is this solving? Is this a matter of addressing public trust? Is this a matter of providing data where there is none before?
Pieter: Right. I think the two things you mentioned, providing trust and providing data that didn’t exist, actually both of them are key factors in what we’re trying to do. The trust-building is increasing. If we look at what’s happening the world around us is increasing a big issue. More and more people are realizing that you can’t trust your own government for everything anymore. I think there is a big value in having checks and balances in our society around data like environmental data. I think in the past, if you would have done this 10 years ago, most people would have trusted their government for collecting data about the environment. Is it safe or not? We know better now that –
Tim: And whether it’s Fukushima in Japan or Flint Michigan in the United States, there’s – yeah.
Pieter: Yes, exactly. There are many examples of that. It doesn’t mean that the government is doing nothing but is the data open available? In many cases it is not. I think what is really happening is that people have a bigger need for knowing what is happening in their direct environments, communities; want to know what is happening. They worry more about pollution and things like that. But at the same time, there is not really a good checks and balances to say that, “Hey, is this data really trustworthy? How can I check this?” Then the second part which you mentioned, is the data available? In many cases it is not. So how do we solve that problem? One is we could wait for our government to have more money to do more things. The other thing is with the current technology, let’s do it ourselves. That’s kind of the Safecast idea.
Tim: One thing I found curious is that Safecast is only available as a kit.
Pieter: Yes.
Tim: Why did you make that decision because, I mean, obviously with something –
Pieter: Yes. Why a kit? I think one thing that we came to know early on, we did a kickstarter in 2011 where we did a Geiger counter, manufacture device. It was very successful, one of the more successful kickstarters at that time. We had it manufactured but took about a year to get it from getting the money collected and getting the people the devices in hand.
If you look at kickstarters, this is a common problem. Manufacturing is hard. It takes much more time to get it right and is very expensive, too. If you make a mistake, it multiplies. We saw that experiences and wait a minute, it’s not really the greatest way to get things out, specifically small series.
If we can make it easy enough for people to solder, there is always somebody in the community who can put it together. It was kind of a bet because we knew manufacturing was kind of out of the question, we also needed money upfront which we didn’t have. We didn’t want to wait a year and do nothing. So we said, “Let’s go and see what happens.”
There were skeptical people who said, “Well, people will never build it.” But it turns out that there are enough people that wanted to do it, and for lots of people became a unique experience to experience building something yourself.
Tim: And I suppose that people who are interested in measuring radiation, a lot of them are fairly comfortable with the soldering iron.
Pieter: Actually not.
Tim: Oh, really?
Pieter: Interestingly enough, not. I would say at least half of the people participating or building have never soldered before. So the first thing to do is either they watch a YouTube video how to solder. In the workshop, the first thing we do is solder your first resistor on the board. We had kids build them, we had people in their seventies build them. It is not impossible.
Tim: So selling the kits, the real advantage is it keeps cost down and allows you to control quality a little better?
Pieter: Yes. It keeps the cost down but most important thing, I think, also if you look at startups and – the cycle time is great because you can make changes to kits any day and the next kit we’ll have to change in there. You also get immediate feedback from people about the device. You can easily adjust for it.
So even though the kit that people build today is more or less been the same for the last few years, actually every couple of months we have a small upgrade to it and we keep on upgrading. The cost of upgrading is almost nothing. It’s literally making – the next board is slightly different or component changes.
If you compare it to 10 years ago, the cost of doing this has gone down dramatically with 3D printers. We use laser cutters here in FabCafe. The tools for us to make that, if you were in manufacturing 10 years ago, that was something that you only could do in a company. Now, that software is open source, that equipment is available in a café today. So that is a huge difference. You can really build devices very easily.
Tim: I guess that makes sense. When you’re talking about a thousand plus devices over five or six years that kind of small batch manufacturing, it’s really expensive.
Pieter: Right, it is expensive. If we don’t manufacture – yeah. Small batch is expensive, and in this case it is a volunteer work so people basically build it themselves so we don’t have a manufacturing cost, we only have a component cost. We just said, “Let’s go give it a shot. See what happens.” It worked out very, very well for us.
Tim: Since everything is kit-based and people are assembling it themselves, how do you maintain data integrity? How do you make sure everything is calibrated correctly all over the world?
Pieter: That’s a fundamental issue. Even if we would use manufactured devices, the same question applies. Just because it is built in a factory, it is also built by people that soldered together. So there’s no difference. First of all, it doesn’t matter. What does matter is how do you know the devices are accurate and accurate over time?
Let’s say in pre-Fukushima, Geiger counters were very expensive because they had to stand on their own. Typically they will be used by people in laboratories or universities or hospitals or nuclear facilities. So these Geiger counters or scintillators tend to be very expensive. And they were calibrated and every year you had to send them in and somebody would check if it’s not working. So enter internet.
We do use a very good sensor; however, sensors can go bad, people may have made a mistake building the thing. So the power of the internet is that we don’t measure one location one time only. The idea is that multiple people that actually don’t know each other will measure the same area over time, and this allows us to see anomalies. If somebody has a bad tube or a bad Geiger, then it will start to show up in our data and we find those and then we will talk to the volunteer. I will take that data and have it fixed. We can do something. So we can see multiple people measuring.
Tim: Right. So as more and more people become part of the Safecast network, this becomes less and less of a problem.
Pieter: Yes, exactly. And even if you spend tons of money on a great Geiger counter, you still don’t have a guarantee that thing is going to work fine. So you’re always going to have to deal uncertainty.
Tim: You guys are doing some pretty cool stuff and you’ve gotten some great press. Are there other groups that are sharing this data or suing this data or contributing to it?
Pieter: Yes. We didn’t talk about it yet but there’s one really important of the project is that everything we do is open and it is published under aCreative Commons zero license. You can go to our website and download all the data we have collected over the last six years which I think is more than 65 million measurements by now. We do hope you give some respect back to us but there is no obligation whatsoever.
Tim: What are some the things people are doing with the data?
Pieter: It’s very varied. The most common use is either local government or local groups that use it to look for hotspots or to confirm for themselves that things are okay or as expected. The second is researchers. There’s lots of different researchers that are doing research about effects of radiation. Then there are the less expected use for the data to the extreme; we have artists that took our data and used our data to make artworks. So we have a Hong Kong artist that used our data to create a virtual tour of Fukushima where you drive in your virtual landscape in Fukushima where the music and the graphics are driven by the measurement data we have for that area. We didn’t know he was doing it because he just found the data, used it and we just found about him later because he was on the web because of this artwork. And we said, “Oh, that’s our data. How awesome that is.”
Tim: That’s kind of the whole dream of Creative Commons, isn’t it?
Pieter: Yes. Joi Ito talks about serendipity and it’s a very important thing. If you think about a startup or if you’re going to think that you’re going to predict everything is going to happen is not going to happen that way. You’re going to stumble. You have to open up yourself for that type of thing. Making things open is not always the right thing to do if you run a company but of course we are nonprofit, we’re explicitly doing that. But it creates an opportunity for people to do stuff with your data you would never have thought about. Instead of trying to envision everything, that stumbling around actually can get you further than when you were constraining yourself with what you think what would happen.
Tim: Yeah, yeah.
Pieter: I think one of the reasons personally why Safecast is still around after six years is that we typically tend not to be too much planning, try to sit down and try to think through everything. We try to experiment and see what works.
Tim: The more people, the more minds that are working on a set of data, the more interesting users you’re going to get for it.
Pieter: Yes. And other funny use is actually is, lots of people love to have an open data they just can play with. They’re not really interested in radiation, they’re just interested in open data that they can play with. So we have all kind of other researchers that use our data to test out their new algorithms and stuff, which is also great because we get something back. They may be able to help us refine our maps or our visualizations over time.
Tim: In an earlier interview, you mentioned that you’d taken criticism from both pro-nuclear and anti-nuclear camps over Safecast. It seems to me that data is data. So what was the basis for their criticisms?
Pieter: I think the polarization that is happening in our society is increasing, which is one of the big things I personally would worry about, has made people so much detached from data or use data in only ‘that fits my purpose’ type of data use has become so prevalent that if you talk about something and it does not fit in whatever your thing is, then you must be on the other side.
Tim: So just the idea that Safecast might produce data that would contradict their beliefs?
Pieter: Yes, yes. Some people had very strong opinions about the radiation levels in Fukushima. There were people that thought that everybody would die in Japan overnight and there were people that said absolutely no problems at all, this is all background radiation. So we had extremes. And these people, they have not looked at the data. They already know the answer and both are wrong from my data point of view. Our point was not to figure out the problems around nuclear power or nuclear energy or anything like that. It is a whole different thing. Our point was, “I want to know what the data is.”
Tim: And we’ll figure out what to do once we understand the data.
Pieter: Yeah. And then based on that, let’s have a discussion.
Tim: The criticism you got, was this simply a very loud minority on both extremes?
Pieter: I think it’s broader than that. I think there were definitely on both of the extremes. But there was a much bigger middle group that had lost trust. So for the same reason they would not trust the government, they wouldn’t trust what Safecast was doing. People, they needed to understand better how we were working. And as we were moving, we spent lots of time explaining to people that this is how it works. We also spent time documenting how we work and we had scientists’ critique. Everything we do, we do pretty much in the open so people can come and challenge and critique and we have lots of heated debates online about ‘is this the right way to measure or is this the right’ – this and that.
Tim: And it’s a truly global community as well.
Pieter: Yeah, exactly. And people come out and say, “Hey, is this really the right way to do it?” So over time, we have been battle-tested in terms of fighting these battles. If anything, we see ourselves as pro-data instead of pro-nuclear or anti-nuclear. That’s the problem we’re trying to solve here.
Tim: Have you largely overcome that trust problem at least here in Japan?
Pieter: I don’t know. I think in Japan, the trust that was lost in the government still needs lots of work. Specifically if you got to Fukushima, there’s still lots of people that have big trust issues where we obviously try to do whatever we could do about that. We have dialogues with government groups, et cetera about what we think they can do better.
Tim: Actually since the Safecast movement, a lot of its roots were in the mistrust of the government data. Has the Japanese government been supportive of your efforts, either on a national or local level?
Pieter: I would say in the beginning nor obstructive nor supportive. We kind of worked in our own space. I think now we have much more interactions where there is – depending on the government. We’re not only interacting with government in Japan. We’re working on a global scale. We’re working different countries. It depends on the country. But there is a general interest to understand how we operate and how this works. The reason for that is this build better trust. It also solves logistical problem. If you look at Fukushima, it’s a huge area. You’re talking about a massive area that got contaminated and how do you measure that? It’s just not possible for a government to have an army ready to go in and measure it.
Tim: Why I could see something. In the exclusion zone is a very special case because that’s not – a lot of places still aren’t safe for people to go back in. But for a city and the rest of Japan, it seems like it would just be ideal to put a bunch of these Safecast devices on the backs of buses and let them monitor the city as they go about their business.
Pieter: Right. Yes. In some cities, we’re actually doing that. They’re not buses but they’re motorbikes. Some cities were doing that. It’s really an experiment to figure out how that works best. Having the technology is one thing but also even if the government would use the Safecast device to measure, it still doesn’t solve the trust issue. You need both. Ideally, you want the government to up their measurements but at the same time you also want to have a way to keep it in balance.
Tim: But it seems to me that that would be an ideal way of getting that trust back because the government is taking their measurements and releasing them, and any citizen can double-check.
Pieter: Yes. Exactly. I don’t think we’re there yet but I think there is more realization that having citizen science – what is the role of citizen science? In this case, citizen measuring, is it useful or not? I think it is very useful. It also is very useful for officials to say, “Hey, we can compare. Are we in the ballpark or not?” If it is in the ballpark, there’s nothing wrong with it, then it gives extra level of confidence on both sides that this is what it is.
Tim: If there is a discrepancy both sides can investigate further.
Pieter: Yes. The other thing is, I think what’s important is how do we make sure that our data which is now becoming our history doesn’t get erased? By having it in two separate entities makes it hard for data to disappear.
Tim: Looking forward, I think it’s easy to see why this is beneficial for everyone involved but Safecast is a disruptive model, and people being disrupted don’t like it. Japan especially, and this is what I’ve heard from so many people who had been on this show, what you’re doing is replacing what has been the territory of universities and government agencies since the end of the war. Did you receive any pushback on that?
Pieter: Not in the sense of pushback in obstructing what we were doing but definitely lots of people were skeptical in the beginning. These are amateurs and blah, blah, blah. If you look at the scientific community, if you have spent you whole life building super expensive equipment to measure and you have a bunch of guys walk in with relatively cheap stuff that actually outperform what you’re doing, of course, you’re walking in somebody else’s terrain.
The reality though is that nobody had a really good answer after Fukushima, and that is, I think, the essence there. Even though let’s say the authorities, et cetera may have had opinions, the reality was they had maybe a few cars that were equipped with equipment to measure at best. They didn’t have the scale or a way to scale up or a way to engage, so they never had thought that through. Of course an accident like this doesn’t happen regularly, and accidents like these are the opportunities to see what works and what doesn’t work. I think that’s where the disruption is. The disruption definitely was quick to react. We were very fast to build and do anything.
In Japan, people are afraid to fail, afraid to take a risk so they would take much more time than we did in building the equipment and test it out. That would be too late. If you’re too late you actually take a bigger risk because now you have missed a whole opportunity.
In Japan, I’ve already been here for a very long time myself, the fear of failure which wasn’t there in Japan after the war, et cetera. It’s something that has built up over the last 20, 30 years. It’s become a major obstructive factor in trying out new stuff. This is a generalization. There’s lots of people that are trying do differently but in general if you look at larger companies, funny thing is nobody came up with a better Geiger counter after the whole accident. They’re still working on it maybe but I’ve not seen it.
Tim: Has the central government and universities come around the Safecast way of thinking over the last six years?
Pieter: I think so. First of all, what’s important to notice is that Keio University was a partner earlier on with Safecast, almost from the beginning. Professor Murai from Keio University, we also worked together with a researcher from Tokyo University very early on. I think in the academic community, there is a very wide range of people but these were the people in the academic community that traditionally didn’t do radiation measurement. They were active in other areas. The people that were traditionally in those places, they took some time to come around. We address that issue more importantly last year by publishing a scientific paper about measurement techniques to address some of the academic discussions where people kept on asking the same question, “Is this scientifically okay?” So we published it in one of the most prestigious journals. We had a peer review; took a year to get a peer review it. It’s now one of the most popular publications in that journal. We are in the top 10.
We went back and said, “Citizen science is after all science. With our volunteers, we worked on it and we published some of the principles behind Safecast and why is this a viable way to measure.”
Tim: It makes perfect sense.
Pieter: And being challenged by the academics is good because dialogue is important.
Tim: Any organization, whether it’s an NP or a startup pursue that same trajectory, when you’re out of the gate, you’re trying anything. You’re taking things together, just trying to solve a problem any way you can and then fix the smaller problems and obtain legitimacy as you grow.
Pieter: If you think about being disruptive, et cetera, I think what’s really important is that if it doesn’t trigger dialogues, if it only triggers people being upset with each other, then you may not be on a winning path. I think creating the dialogues and being critiqued and also taking in critique while you disrupt and being open to that and of course quickly react to that actually is the way to go.
I think over time people have come around and say, “Well, this is a whole different way of doing things.” So we have inspired lots of people in other areas, not just in radiation measurement but in lots of other areas where they try to build a similar type of model for solving their problem.
Tim: I find it incredibly encouraging that Japan in particular is taking to this philosophy as well as they are, the idea of citizen science, of data sharing, of open data. How do you think citizen science is going to develop in Japan over the next 10 years for example? What changes are we going to see here?
Pieter: I don’t know. Because citizen science, it’s like a biz word. We have been doing so-called citizen science for six years but I’ve seen other groups doing things in very different ways. What I’m sure about though is that we’re going to see more citizen science-like projects that are increasingly more successful and more impactful. The reason for it is very simple. It’s that technology that drives it is getting cheaper and cheaper. The access people have to it is getting better and people are having better tools to corroborate on that. What we have seen is that we have inspired quite a few researchers or other groups doing the same thing with something else. We’re teaching people how to do Safecast. We’re sharing our way of thinking so I’m quite sure more of this will happen.
Tim: What other kind of problems are people attacking with citizen science?
Pieter: Okay. One example is a researcher that is trying to focus on the quality of water. He’s been trying to figure out how can we measure water at people’s homes with the same thing. He wants to do it similar way as we’re doing. Another researcher has been trying to measure the ocean for radiation. Ocean is very big as you know and in order – there’s a technical problem with that. You can’t just put a Geiger counter above water and measure it. You need to actually take 20 liters of water, boil it until it is completely evaporated. And then the residue, you put in a Geiger counter; you measure it. It’s not so easy. He was doing this on his own as a researcher, and he’s one of the leading researchers and he could only get so many samples as he could cook water.
Tim: I could imagine, yes.
Pieter: Right. So just imagine cooking seawater. It’s not entirely a great experience. What he has done is, after we met him, he said, “Wow. This is a great idea if I can get people to collect the water for me.” Se we teamed up and he started to do the same thing so now he has a kit that he mails to people. They fill it up with water. They send it back to him. He does the cooking and measurement. He sends back the results to the people that collected it and he gets the data he’s looking for.
So there are many inspirational things that are spinoffs of the same idea and I’m quite sure there will be much more of this. I think in the whole developing side of the world, there will be much more of this than there will be in the developing side. If you look at places like Africa or Asia or whatever where there is huge pollution problems, et cetera, I think people are going to be armed with much cheaper equipment that will allow them in their communities to have much more impact about what is happening. I think that’s where we’ll see citizen science revolution.
Tim: Excellent. I hope we do see that. Listen, before we wrap up, I want to ask you what I call my magic wand question. You spent six years trying to get this innovation to become mainstream in Japan, if I gave you a magic wand and I told you you could change one thing about Japan, anything at all, the legal system, the education system, the way people think about risk, anything at all to make it easier for new innovation to take root in Japan, what would you change?
Pieter: I think it’s a combination of the educational system and the attitude towards risk. I think that’s a big thing. If Japan can change that, they will have a much better way to innovate. So basically, the educational system, teaching from kids onwards about failure can be a way to learn, failure is not necessarily bad, failure can be very good. People are way too much worried about the downside than the upside. Failure by definition is not good or bad. But definitely making failure by definition bad is a big handicap. I think it roots in the educational system, it roots in some cultural – but if you go back in history, not so long ago the young entrepreneurs after the Second World War were very different from the people we see right now. They weren’t afraid of taking a risk or failure. They were actually quite courageous. I think things have somehow gone into a direction that values low risk over progress. If that changes for innovation in Japan, I think it will have an impact.
Tim: All right. Listen, Pieter, thank you so much for sitting down with me. I really appreciate it.
Pieter: Yes, thank you.
And we’re back.
We’re not quite there yet but projects like Safecast and citizen science, more generally have the potential to invert the relationship between government scientists and citizens. You see, environmental monitoring and research projects are usually far too large for an individual organization to undertake, and historically the solution has been to have the government take on that responsibility on behalf of all parties.
It’s a good system and it’s worked pretty well so far. What’s interesting here is that now some projects are simply too large for the government to undertake but individual concerned citizens acting together can do the research and get the data. There is no way a government program would be able to get funding to monitor radiation or pollution on every corner of every street. But concerned individuals working together are already heading in that direction and quite frankly, anything that involves citizens collaborating directly with a government agency towards a common goal or citizens working directly with a scientific community will go a long way to improving peoples’ understanding of government and science. And it likely will improve the quality and responsiveness of both government and science.
The fact that Pieter and the Safecast team are now winning over the scientific establishments on their own terms by publishing in peer-reviewed journals is something that all startups can learn from. This is the other side of disruption and it’s the more important and more difficult side. The real value in disruption is not the destruction of the old way of doing something. The value is replacing it with something sustainable and better, with something that improves peoples’ lives, something that becomes the new way of doing things.
If you’ve got an experience in crowdsourcing or citizen science, Pieter and I would love to hear from you so come by disruptingjapan.com/show089 and let us know what you think. When you come by the site, you’ll see all the links and notes that Pieter and I talked about and much, much more in the resources section of the post.
But most of all, thanks for listening and thank you for letting people interested in Japanese startups know about the show.
I’m Tim Romero and thanks for listening to Disrupting Japan.