#25 - Achim Haug & Anika Krause: Open-Source Air Quality Monitoring and Community Engagement

Achim Haug and Anika Krause

Achim is an ex senior Seimans exec of some two decades with a masters in science who now lives in Thialand and founded Air Gradient, an air quality monitor with a unique approach in the sector. 

That started as a volunteer project to help a school in Northern Thailand monitor the air quality in classrooms during the highly polluted “burning season”. 

From the beginning he put a strong focus on open designs, robust and long-lasting hardware and a strong desire to support people from all around the world with affordable and effective air quality solutions. 

Anika is Air Gradients scientific advisor and lives in Italy. With a  Masters in science and a PhD in philosophy and atmospheric sciences from Cambridge, an expert in air pollution exposure science and data analysis. She has spent several years in research and industry Before joining Air Grandient last year.

AirGradient is currently the only global air quality monitoring company that makes all its hardware design available for free under a Creative Commons CC-BY-SA license. This means that all electronic schematics, firmware code, CAD files are openly available and allow (and encourage) other organizations to use them. They openly publish all their research and algorithms so that others can benefit from the extensive work they are doing.

Its a fascinating approach that cuts against the mainstream and opens up all sorts of potential, risk, challenges and benefits.

At a time when the air quality community and particularly the low cost sensor sector is figuring out the what next, Air Grandient is forging a unique path. And one I was fascinated to discuss.

Achim Haug - Linkedin
Anika Krause - Linkedin
AirGradient 

Transcript

Simon: 0:14

Welcome to Air Quality Matters, and this is a conversation with Akeem Haag and Annika Kraus. And Annika Kraus. Akeem is an ex-senior Siemens exec of some two decades, with a Masters in Science, who now lives in Thailand and founded AirGradient, an air quality monitor with a unique approach in the sector. It started as a volunteer project to help a school in northern Thailand monitor the air quality in classrooms during the highly polluted burning season. From the beginning, he put a strong focus on open designs, robust and long-lasting hardware and a strong desire to support people from all around the world with affordable and effective air quality solutions. Annika is AirGradient's scientific advisor and lives in Italy. A master's in science and PhD in philosophy and atmospherics from Cambridge, she is an expert in air quality exposure science and data analysis. She spent several years in research and industry before joining AirGradient last year. Airgradient is currently the only global air quality monitor company that makes all of its hardware design available for free under the Creative Commons license. This means that all electronic schematics, firmware, code and CAD files are openly available and allow and, in fact, encourage, other organizations to use them. They openly publish all their research and algorithms so that others can benefit from the extensive work they are doing. It's a fascinating approach that cuts against the mainstream and opens up all sorts of potential risk, challenges and benefits. At a time when the air quality community, and particularly the low-cost sensor sector, is figuring out the what next? The low-cost sensor sector is figuring out the what next? Airgradient is forging a unique path and one I was fascinated to discuss. Thanks for taking the time to listen.

Simon: 2:34

This is Air Quality Matters and a conversation with Akeem and Annika. Look, it's really great to have you both here For listeners. This is the first time I've had two guests on the podcast at the same time. Not only that, we're dialing in from Thailand and from Italy. We've got beeping, which I'm assuming is Italy is it or my pigeonholing Italian driving and beeping of horns, or is that where you are, akeem, outside, and we've got thunderstorms as well going on where Akeem is, so there could be all sorts of fun and games in the background with noises, so let's see how we get on.

Simon: 3:19

The question I wanted to really ask you both and it was one of the principal reasons for wanting to chat chat was that I've spoken to you, akeem, particularly about this sense that feels like this precipice of wanting to change gear with air quality monitoring, that the built environment has gone through this change, particularly through the COVID period of recognizing the need to manage the indoor air, and that somehow we need to do more. And we've come out the other side of that and I think it just feels like it's time to shift forward somehow, and I wonder if it's time also to start thinking a little bit more outside of the box, start thinking about a more open approach to how we monitor the indoor environment and perhaps even a more open source type approach, and I know that's something that both you, akima and anika, have spent a lot of time thinking about yeah, um, simon, I think it has a lot to do like how we actually started at Gradient.

Achim: 4:26

So I was on a sabbatical in Northern Thailand and they have a fifth season they call Burning Season and in 2019, it was really, really bad. And I was volunteering at my children's school as parents' president at my children's school, as a parent's president, and we could really see how the school struggled to understand how good is the indoor air quality? Is it safe inside? Do the air purifiers work? It's a huge school. It's a boarding school. It's like 100 acres. The head of school was running around with a meter trying to make sense of the situation, running around with a meter trying to make sense of the situation. And so me and my co-founder, we saw that and said, look, there has to be some kind of better system in place, because we saw how much it took away from teaching, from just basically the whole school day. And with that start, where we also then looked at the underlying reasons for that smoke crisis in Thailand and started working with the community to understand, hey, what can we actually do against it? What can we actually do against it? We started then building monitors and actually also like air purifiers, together with students and parents, and so this kind of openness that is basically defining us today started on day one with where we came from and another thing I think that shaped, let's say, from a company perspective, the openness, was that, being located in the global south, you know where the income levels are just a fraction of what we have in the west, and also there are such many more pressing issues. You know, like people still concerned about clean water, for example, you know that there's just not this kind of scope that they can go and go out and buy an air quality monitor for like five or six hundred dollars. It's just not working. It's just not working. So, just kind of being in this region here we also saw, hey, there's really a need to have much more affordable air quality monitors. And also we saw it as a responsibility, because we started as a volunteer project.

Achim: 7:00

We didn't start with a business plan to say, hey, let's maximize the profit, how can we actually make an impact? I think that was really there from day one. We need to help people protect themselves from this hazardous air pollution. We have to enable people that don't have much money to protect themselves. How can we do this? And that's where, basically, this openness started, where we, from day one, when we first developed we call it DIY kits at that time Basically very simple build instructions how to build an air quality monitor we put it on the internet and just said, okay, it's there so people can use it. At that time we didn't monetize it at all. We didn't even sell any kits. What we do today. We just thought, okay, let's have it there Because we felt it's good to share it and to enable people.

Achim: 8:05

And then what happened next was that, well, that volunteer project, we got more and more attention. So other schools came to us and then people called us from Bangkok and then people called us from China and we thought, ok, that's interesting, because China had such a head start in air quality management, because China had such a head start in air quality management and if they see some value in what we do, you know, I mean in the meantime we built up also like a dashboard application and started like developing, let's say, more professional monitors. We felt okay, like especially for me. It was like, should I go back to my old job or not? And it was just very clear that you know my passion. Is there Also the feeling that I can have much more impact in that area. And only then, a couple of months after we started, proponents now as a company is really in the DNA from day one. Okay. So it's not so much that we felt at the time when we went open that we felt, oh, this is really really a need to make money. We saw more. No, this is actually our way to maximize the impact and the let's say how we can help and support people also in the global south, how we can help and support people also in the global south.

Achim: 9:49

So what was interesting then was that when we developed the first version of our indoor monitor at GradientOne, we kind of started or we thought, okay, we have to imitate, like the other players in the market, and so the first version was actually closed source. And then we came more and more into like, let's say, other customer segments that like maybe valued more like I don't know how to put it like this notion of like I have of like I have to like certify a building just to get a label and tick. I mean that was also just like during COVID or maybe a little bit after COVID, and we felt more and more that this is actually not exactly the space where we want to be as a company, because it was also this places, places where, okay, like you put maybe one monitor on one floor, you know, um, but we, we know that every, every room can have very, very different air quality, but that was often not really covered in these standards. Um, and so very, very quickly, we actually made the decision and said we actually want to not follow the other companies but really go back to our roots and say, hey, let's try and be a completely open company.

Achim: 11:18

And then something interesting happened. So we made the decision to be completely open source hardware. So we made the decision to be completely open source hardware, so that means like, if you go on our website, you can actually download the firmware from the monitors, you can download the enclosure files, you can download the schematics, the keycat files, you can build this whole thing, the whole monitor, yourself, and we even allow it to you with a Creative Commons license. So this was something that I think still today no other manufacturer is doing, but for us it was really the logical consequence from how we started. I think it's super important that it reflects the core beliefs of the founders and then of the whole team as well. So we basically just took a risk in a way, because I mean, there was nobody in the industry that we could say, ok, let's follow them, and said, ok, we want to get become really, really completely open.

Achim: 12:25

And and then interesting things happened, because by being open you actually get a lot of attention. First because we were like the let's say that the first ones doing it, but you also enable a lot of new customer segments or new applications. So one of the first people we got really interested in was actually the research community, because often they have the problem, like when they buy air quality monitors for their scientific studies. They want to know what's going on in the monitor. They want to know do they already apply certain correction factors? For example, how often do they measure, what is the frequency, how do they average the data? And a lot of companies regard that as their trade secret. So often the researchers have really problems like, for example, if they work with different monitors, how they can really compare the data as being like open source, they can just read the code. And not only can they read the code, they could change the code and say, hey look, we need a higher frequency and then basically put that back on onto the model.

Simon: 13:49

So, um, yeah, this is where how we came to that, let's say, open approach when we talk about the advantages that something like a more open approach can have, it seems to me that there's several elements to that, or several legs to that stool. There's the obvious advantage of being open about your hardware and your technology and the firmware and the software that runs it and that can do things like, uh, open up transparency conversations and enable academics to understand how and why they're seeing the information that they're seeing, or enable certain communities or organizations to develop certain elements off of that platform. So there's that flexibility of the hardware and that transparency about what it's doing. But I'm guessing there's also that open approach also brings in a citizen science element to this as well, that it's not just about the hardware and the platform. It's about being more open with the kind of data that we're collecting and how you do that, and engaging with stakeholders in a different or more meaningful way I think that's one.

Achim: 15:07

One very important aspect is like, for example, when, when, when I talk to potential customers, by being completely open source hardware you, you kind of take the sales pitch out of it because I can immediately say, hey, you can either get it from us or you can build it yourself. So this just puts the whole conversation immediately on a different level. So that also builds, I think, a lot of trust as well, because by definition and I used to have that actually as an email signature for some time I had written like we do not have trade secrets as a company. So by being open you don't have secrets and that's automatically building trust and I think that's a super important element when you engage with potential customers. But that's only one aspect of the openness.

Achim: 16:07

I think there are a lot of additional benefits like, for example, it's easy to, you know, extend the hardware. You know like often people want to add I don't know just one more sensor. It's maybe pretty simple to add and because it's an open system, it's easy to do. Other things are just like it's easy to repair. You know it's like I think from iFixit we would get 10 out of 10 points. You know the schematics are there, people can repair it and it extends the lifetime of the monitor. And I think that's also then coming into this affordability thing, like if, if something is easy to repair and, um, it could run for 10 or 15 years, you know, whereas other products you maybe replace after a couple of years. So I think there are a lot of benefits. Maybe Annika has some other ideas to add here.

Anika: 17:15

Yes, because Simon just also mentioned citizen science and perhaps I can add that, well, our open source approach is, on one hand, about open hardware, which can trigger affordability in the end, as Achim just says, which is helpful for citizen science. But we also have other aspects of openness For example, that we share our data, like the outdoor data, that we share our data or the outdoor data. Everyone who has a monitor, an outdoor monitor, can just with one click, share this data with the world, basically with scientists, with the government. So that is also really useful if you're a citizen scientist, so we have this kind of infrastructure to share data with research, to make some sense out of it. And also, we are aiming to not only provide data like in different ways, making this possible that there's more access to air quality data but also we're working towards helping people to understand them, to interpret them, so we we provide trainings on what does that data actually tell me and also what can we do about it.

Anika: 18:33

So, for example, simon, you mentioned also that the built environment, the air quality in built environments, is a problem, and it is Only household air pollution and that's not schools or offices including is causing two and a half million premature deaths in the world every year. And well, the first step is to to address the problem by understanding it fully, measuring indoors, but it's also important to then make something out of this information. So there's also a need to um, to increase the awareness of people, and to to give them an understanding what what they can actually do about it, and this is also one of our missions, basically to, uh, yeah, to educate people on that and provide this knowledge as well as a kind of open source approach.

Simon: 19:27

I think it's a really interesting point, annika and it was something I was just having a conversation with somebody about this morning that we've started to go on this journey in the sustainability world of transparency, particularly when it comes to things like non-financial reporting, that there's a rigor being applied to sustainability. That means greenwashing and vanilla sustainability reports don't really cut it anymore. That everybody recognizes they've got a long journey to go on and the more open we are about that and the more benchmarks and standards that we can lean on for support, the better, because it enables us all to make the right decisions. And I wonder if there are lessons from that world that we can be learning from the air quality community that at the moment we seem to be at the early days of sustainability. The vanilla uh, healthy building reports and the good intentions to be a healthy housing provider and all of those. They're all kind of where sustainability was perhaps a few years ago. But but we've seen sustainability go on that journey where you need more and more measurements and metrics. You need to back up what you're saying and I wonder if that's perhaps a journey and we've seen it with the outdoor air community as well that's much more open source.

Simon: 20:57

The challenge is this fear, I think, of the built environment, of opening up your building for scrutiny and underperformance and risk, and, and it's I wonder how you cross that threshold. It's, it might be different in public buildings, in schools and in academia, for research, but when you start crossing the thresholds of people's homes that are being provided under social housing or private businesses, um, it becomes a sensitive subject, doesn't it? Openness in those circumstances, are you getting into those sectors with what you're doing it? You know, I'm guessing academic studies and those kind of things, yes, absolutely. But are you seeing that same interest from the private sector as well for this kind of approach?

Achim: 21:48

I think, um, yeah, it's important to separate, like indoor and outdoor air quality here, and let's maybe talk a little bit about indoor, because probably more more interesting and more challenging aspect we have a lot of conversations about actually also crowdsourcing indoor air quality for research, and then one of the first things that comes up is really, can we anonymize the data? And open data does not necessarily mean it has to be personally identifiable. So what? Because there were some really interesting aspects. For example, I think in some cities in the US, when the city provides social housing, the occupants go into a contract and say we don't smoke in that apartment. Now, if you would go in and measure air quality there, I think it would be pretty easy to detect smokers, for example, and then these people could potentially get fired out of their housing contract and, being a social housing, they might end up on the streets and that's obviously not something we want to enable as a company. So, on the other hand, you could anonymize the data and then still provide insights into how these buildings perform to enable improvements.

Achim: 23:13

And I think another really important aspect is especially for indoor data, and it's also something we are working on is you need a lot more metadata than for outdoors. On is you need a lot more metadata than for outdoors. So we are actually in the process to implement the kind of questionnaire inside our dashboard that people can, for indoor monitors can opt in to share the data with science, similar to what's already done with outdoor monitors, but they would need to fill in a question, for example okay, what kind of ventilation system does the house have for that room and what kind of floor is it? Um, you know, is it carpet or is it stone or wood or whatever? Do you have?

Anika: 23:59

a gas stove for an electrical stove in the kitchen. These kind of things.

Achim: 24:04

How many people are in the room? Um, are there smokers or not? I mean, that's maybe a bit of a critical question, but um, so like, by collecting this metadata. I think this is like essential to really understand the data, but then like um the the idea behind that project is to really crowdsource indoor data, similar to what citizen science do on the outdoor side already since many, many years, to enable that with indoor air quality data, and we see demand from research to to get these kind of data sets I, I know research is always hungry and you know, and I, I, I get the sense they feel the built environment is slipping away from them from a data perspective, because the business cases and the commercial argument is getting stronger and people are deploying sensors and data and hardware and academics have no access to any of that.

Simon: 25:14

There are billions upon billions of data points being collected in all sorts of really useful environments. That is an academic's dream, but they have no access to it and it's and it's interesting to me how we bridge that, how we cross that threshold. There's a lot of discussion in in the uk, as around indoor air quality observatories, you know, about setting up organizations that are there to coordinate these approaches and create places where data could perhaps be reposited and used for analysis by academia and others. I think it's a fascinating area because there is so much data being collected now from different sources that it's a missed opportunity if we don't find ways to bring that together for learning, for sure.

Achim: 26:12

But what I often wonder, simon, there is that, yes, there's so much data collected and it's a shame because it's not available for research, but often at the same time, I think, is there actually any value connected to that data? You know, I mean, it seems to me in a way that, yes, a lot of these big companies collect all the data and then it's somewhere locked away and they maybe even themselves don't really know what to do with it. You know, and, yeah, it's just a shame, I think don't really know what to do with it. It's just a shame. It reminds me of an anecdote when you go on our website, you don't see any of these cookie banners, these GDPR compliance banners, because we just stopped collecting any personal identifiable information with our web tracking. So we got rid of Google Analytics and took some more privacy-friendly tracker and we don't miss anything.

Achim: 27:13

So before we used Google Analytics, it collected tons of data. We never used it as a company. And then, back in January, we made that decision why collect this stuff? If we don't need it? Let's just stop collecting it. Our website became three times faster. Um and um. It's a privacy win for for the customer. So that's what, what I think is often really this kind of companies saying, okay, let's just collect as much as we can because we can. But the question is, do you really need to, or could it be made much more useful if it would be shared?

Simon: 28:00

basically, yeah, I have a saying that, uh, data without purpose is just noise and it it. It comes from my perspective as a commercial person, often looking to understand why we would be doing something and why somebody would value something. And you have to ask the what next, what next? Questions. You have to root cause where the value comes from, and we don't do that enough. A lot of the standards and a lot of the hardware we have is a because we can approach and we'll figure out what to do with it downstream and that very rarely manifests into something valuable.

Simon: 28:41

I had a really interesting conversation on the podcast with Tom Robbins from Switchy, who are a social housing environmental monitor, smart thermostat provider, and they very much come from a social purpose, a bit like yourselves background.

Simon: 28:56

Why do we exist?

Simon: 28:57

Well, we exist to make the lives of people in rented homes better, so every bit of the data that they use is used for that purpose and therefore building the business case to social landlords is easier, and I think we've got a lot of that to do in the built environment.

Simon: 29:17

You know, know, schools, which I know you operate in, is a very good example of that. You know, we saw a proliferation of sensors into school classrooms during the pandemic, most of which are now in drawers or turned around because the kids keep pointing at them and it's annoying for the teacher at them and it's annoying for the teacher. We never really asked ourselves the question of what the value was of those beyond the COVID pandemic and what the organizational change management that might be required to get better outcomes from that data, or how you would engage with all the different stakeholders in an educational setting, from kids through to teachers to year heads, to school managers. All of them have to be involved to get a better outcome. How do we do that in a meaningful way? And I think that coming in from an open approach and asking the question why are we doing this is a good starting point, because you're starting from the right perspective and then you can build your case around it openly.

Anika: 30:23

I think, even if it's not shared with the whole world, with researchers, but the occupants at least of this building are informed, that would already help a little bit, like I. I mean, yeah, you're saying the question is, why are we collecting this data? And if it's just to raise the awareness of the occupants and they then understand when to open the window or when to not open the window, that is already, I think, an achievement, like it's another purpose, it's not driving research for further, but it is changing behavioral to protect yourself, like your behavior to protect yourself from harmful effects of air pollution. So I think it is valid, like if the occupants see these data, if you just measure it and save it somewhere on a server and no one looks at this, then I agree with you, of course, that that's a little bit pointless, but it can also be used for behavioral changes without you know this step in between of research and do you?

Simon: 31:28

do you find people practically leaning on this openness that you're providing? So are people taking your products in a meaningful way and adding to them, or are you having regular conversations with people? Again, yes, I'm using this as an idea, but I'm going to do this with it instead. Could we have some help with that in practice? Are people taking that open source nature and doing stuff with it.

Achim: 31:57

Yeah, a lot. So, besides the science community, we have a really big we call it tech crowd. That's maybe more than half of our sales and they are just people that want to integrate the monitor within their smart home system, for example, and there are different smart home systems, like smart home system, for example, and there are different smart home systems. So they're actually contributing like integrations to like Home Assistant or OpenHab or all these things that are out there, and then they just basically commit the code on our GitHub repository and we merge it into the main code. And that's one aspect. So, basically, instead of us hiring our own developers, I mean we have core developers, because you cannot run it completely with volunteers. But a lot of extensions are getting contributed already and we have also a very, very active forum, um, where people really discuss not only technical issues. But we had a super interesting um discussion about vocs and modern sensors and that they actually monitor the changes of vocs rather than absolute values, and I think a lot of actually established players don't get that communication right. Um, and there are actually mistakes in some monitors which still shows ppb value when it's actually an index of of a change and um. I think that the community also really improves the quality of the product. You know, so, um, we, um, I think one and a half years ago we had the problem that the monitor was running not stable, so it was like, but it was not crashing very, very, very often too, but it was just like randomly rebooting and we didn't really know what's going on and we put a lot of time into like trying to figure it out and we, we just didn't know it.

Achim: 34:05

And then on on the forum, people started discussing it and then there was one guy called Ken and he started writing longer and longer and longer posts and really digging into spec sheets and technical data and you know, like publishing or putting screenshots from his oscillograph. And in the end he found out what was the underlying reason. It was actually in a library that we used for I2C communication, in a library that we used for I2C communication. And then I went on a call because I said, hey, ken, let's have a chat, because I was super curious what's his background. And then it turned out that he's based in Silicon Valley and develops hardware for artificial intelligent computers, so he probably earns I don't know a couple of thousands, $100,000 a year. And just because it's an open source project and he got interest in air quality. He was just helping us with his expertise to find that back. So there's a lot of contribution from the community and it's only possible because it's open source.

Simon: 35:21

So you're really leaning into the Linux Reddit geek forum enthusiasts, diy engineered. There's a massive, massive community out there of technical people that build forums and get interested in this stuff, and I've not heard of that in air quality before. So that's fascinating to see how that will develop over time.

Anika: 35:48

I think another aspect where our openness resonates is this openly sharing of data and with a with a community, with the world, basically. So we had people contacting us and saying like I'm sorry if I accidentally pressed not share my data, like refuse everything, or something like that he said, because I usually do it because I'm concerned about my privacy, but I trust you guys and I want to actually support you, so how can I share my data with you? And we actually had a couple of people who asked like how can we do that? So now we published it a bit more like instructions how to do that. So people stopped because it's easier to find out now. But yeah, so we actually had people coming towards us and asking us how they can support us in the data sharing. So it's kind of feels maybe almost like a movement, that you're part of a bigger thing, so and and a really nice community. So yeah, I think that's another nice aspect of of this open data sharing it'd be interesting to see where that goes.

Simon: 37:00

You know where ultimately that leads. Do you find that distracting? Though, at the same time? I mean, you are running a business and trying to develop products and you know you can't please all of the people all of the time, as they say. All of the people all of the time as they say, um, I imagine, as interesting as it is and as valuable as it is, you could probably find yourself getting sucked down black holes and falling down rabbit holes of uh, technical, you know discussions.

Anika: 37:31

So far, I would say I find it very encouraging, like it's nice to have this feedback and to hear that people are interested in you know, achieving the same impact or goals that we want to.

Achim: 37:43

So I can imagine that at some point, if it becomes too many people, it will be hard to answer everyone personally, but so far we manage and it's I find very encouraging yeah yeah, maybe to add like so far, maybe the good thing is that you know air quality monitoring still a very, very defined area, unlike you do in linux, linux project or something, and it's also you have quite a lot of limitations when it comes to, let's say, well-performing sensor modules that you could actually put into the monitor. And we had a couple of discussions, like things came up a few times and then we discussed it and actually the community also discovers okay, actually that's not easy to implement, you know, or maybe that's not an accurate sensor module, so maybe it doesn't make sense to do it. So I think the key is really to to very, very openly and honestly and also having a science-based communication with the community. And then you know, and then in the end it becomes easy because if after that conversation a certain feature makes sense, then okay, why not implement it? Or it comes to the conclusion that it doesn't make sense.

Achim: 39:13

Or there could be a case, and let's make one example. So a few times, people want their power over Ethernet, as you know, just getting the power to the monitor, and that's something that's technically very, very tricky with the microcontroller we use and it would make the product a lot more expensive and then actually the community members made their own design and shared it on our forum. So it's not necessary that we have to implement what the community wants. That would be more like this kind of traditional company approach where the customers say, oh, we need this and this and this, and then the company decides. And then the company decides. But because we have an open source hardware design, people can just basically what they call fork in computer language, just take the existing design and edit and let it run as a parallel project and no problem. And then if people need that feature then they can go there and get it implemented and and and get it implemented.

Simon: 40:23

And do you find that helps with um starting to understand the boundary conditions of the potential of these kind of sensors and the boundary conditions of the operating models they might go into or the kind of sensing conditions that they might go into? Because there's a it's. All sensors have their limitations and by having that kind of multiple stakeholder open approach to it you can really test those boundary conditions quickly and fail quickly. People, somebody can take a notion to an irish saying a notion, but take a, an idea to go down a certain track. They think there there's value in using, let's say, a metal oxide sensor calibrated to measure formaldehyde.

Simon: 41:03

I think I've got a use case for that and they run off with it for a bit and come back six months later having chatted to a bunch of people in the forum and go no dead end for these reasons and these readers. And then you haven't spent six weeks investing time and money trying to demonstrate a business idea, because I've been on the other side of that in with manufacturers as well, as a lot of time and effort goes into trying to work out whether there's a business case for something. Everybody always wants everything, but at some point as a company you've got to decide where the business case makes sense. If you've got a whole community doing that work for you in a way coming up with these ideas, testing them, see, finding the limitations you know that must be an interesting position to be in.

Achim: 41:52

Yes, I think there are like two aspects there. One is like let's talk about modern voc sensors you know, like from since the end. For example, there are a lot of tuning parameters you can do, which hardly any company is doing. They just take the defaults and it's also probably very, very difficult to tune them. Or you need to know, like what you just mentioned, you, you need to have a very, very specific business case, which we, for example, as a more general, you know kind of company that offers the monitors, don't really want to go into a very, very specific use case. But then, because it's an open source code base, you know somebody could very, very easily change these parameters, flash them onto the monitor and try these uses.

Anika: 42:40

If you really want to be like a researcher and you want to have a thorough validation of your sensor, you need to put it next to a reference instrument and have it, you know, expose it to all kinds of different conditions, different emission sources, different temperatures, relative humidities and so on.

Anika: 42:57

So basically, you would actually have to travel around the world with a huge, expensive reference instrument and do it everywhere.

Anika: 43:05

So this is pretty much impossible actually for any manufacturer to do. However, we have this community, which is also a lot of academics and governmental institutions are in there or big NGOs who possess these reference instruments all across the world, and since they value our openness and there's trust and collaboration between us, they let us use these reference instruments. They let us, you know, use these reference instruments. So basically, we send them some of our sensors and we can use this synergy of our community basically to properly validate and on a scientific level, basically our monitors. So this is also a really nice point about our community and data sharing. Also good is that, you know, working with a lot of universities together, we get a lot of feedback also on how to like ideas on how to improve the performance of our sensors and or what other experiments we could do or what other sensors might be, you know, have the best performance if we want to integrate new sensors, and so on. So this is also a really valuable aspect of our openness and community sharing aspects and minds.

Simon: 44:27

Are you able to kind of create a structure for the less academic out there to help them assess your product better, because there's an approach to science rigor that academics are hopefully following that you can rely on when they're testing it against reference instruments and so on and processes. So that's all very valuable, but I imagine there's a tier or two down from that of people that have a lot of technical competency in different areas and want to test the boundary conditions in different scenarios, that if you can provide the right framework for them in this kind of citizen science approach, you can get value from that tier. More effect you know how are you labeling the data, what kind of frequency? You know, can we all agree on normative values for things, so that we're measuring apples with apples and that you know that those kind of approaches it's a bit that.

Anika: 45:20

I think that's really, yeah, interesting idea, annika yes, we do actually, um also want to make sure that not only academics understand this validation and this data, and we also want to make it as accessible as possible for a wide audience, like everyone who was interested should be able to understand this data.

Anika: 45:41

So therefore, we also give webinars explaining on, for example, like how to quantify an error of a sensor and things like that, so that and we explain it really, you know, step by step and not with not too complicated equations and so on, so that everyone who's interested can actually do the same. And while private people usually don't have a reference instrument for half a million dollars standing in their backyard often or it depends on the city but often you have air quality networks which also publish data, so you can try to get your sensor close to that or compare it. It doesn't have to be like just one meter away. You can still get an idea of how well the sensor performs when there is a reference station in the area. So you're right, simon, it's important also not to have the academic world understanding all this data, but we want to make it accessible to everyone and therefore we try to also provide the tools to interpret these data maybe explain to people what it is in its standard form yeah, so we, we have two monitors, one indoor, one outdoor.

Achim: 47:02

They're both sub-200 US dollars and we sell them in two versions. So one is a fully assembled, certified product the same. You can get it at Purple Air or Kiterra or AirThings. Exactly the same product. You plug it in, you connect it to your Wi-Fi, you have a dashboard, you can see the CO2, the PM. Exactly the same.

Achim: 47:27

So we have a lot of users, obviously that are doing that, and we also got there's this micro-sensor challenge from AirLabs every two years, and last year actually we got three awards and one was the most accurate multi-pollutant sensor under 500 euros.

Achim: 47:45

Even so, we had $200.

Achim: 47:47

So you get a very, very accurate monitor using pretty much the same sensor modules like everybody else, but at the same time all the design files are open, the firmware is open.

Achim: 48:05

So for the people that want to go beyond just using the standard product they have, then basically I mean all avenues open so they can open the monitor with four screws and put an additional sensor module in there, and there are basically pins that they can easily connect it to and then download the source code, put in a little bit of code to get the data out of that sensor. And then it continues like they can send the data to our data platform, but that's totally optional. They can also set up their own server and change the code and send the data directly to their own server. So it's really, basically and that's the beauty of the whole thing it looks like all the other monitors out there when you buy it, but it's just a completely open product. So we don't lock it down, we don't push people onto a certain platform. It's just we give that total freedom to do whatever they want with it.

Simon: 49:16

And, like everybody else, you're not making the sensor sensors, so I'm guessing you're using are you using typically the kind of the bigger main brand sensor supplies the Sensirians, Sensairs, EEs, Bosches, those types of products or are you innovating and using alternative sources for the standard kind of sensors that people would expect to see in those kind of products?

Achim: 49:47

Yeah, we use the same third-party sensor modules like everybody else. So we use the Sensor S8 and the IR CO2 sensor, for example, which is a very well-performing one. For temperature and VOCs, we use sensorium sensors and then the plant tower module for, like, pm. So that's a very, very common set of sensor modules. So we don't manufacture or design our own sensor modules, but look at basically what's the best available low-cost sensor modules in the market and use them.

Simon: 50:32

And I'm guessing you're getting a lot of rigor applied to that. With your open technical community, they'll have an opinion very quickly on one module versus another, I'd imagine. Um, so I'm so I'm guessing, then you did you're doing particulate matter via tvocs temperature, humidity, co2 as the kind of the. The core grouping are you, we?

Anika: 50:56

also have noxious, but this is more sorry. Yeah, we had a connection issue.

Simon: 51:01

Yeah, you said uh, you're also looking at the noxes. I was going to ask that question are you straying into things like nox, ozone, formaldehyde, some of the other things that we're starting to see being measured?

Anika: 51:15

yes, so we currently have a metal oxide sensor from OXIS, but this is more looking at relative changes rather than absolute concentrations. But we do see that there is a demand for NO2 sensors, like specifically NO2. So that's what I'm actually currently doing looking into different sensor models and perhaps testing them a little bit to see which one is performing best. And yeah, that's one route we're planning to go down, like to integrate this, and usually the best sensors are electrochemical for that. And so I don't know, I haven't really like we haven't really decided on that with, so it's nothing defined. But once you have NO2, it's quite, I think, also pretty easy to have ozone in it. Like it's the same technology, so all electrochemical sensors are kind of interchangeable. But, yeah, I am not so much into the technology behind it. So, yeah, I don't know if it's true.

Simon: 52:20

yeah, I think everybody's pretty comfortable with both the potential and the limitations of that kind of main set of four or five sensors that we're using. Yes, some people missell their capabilities and other people are looking at them in the wrong way, but generally speaking, we know what we're getting at this stage with metal oxide sensors and NDIR sensors or photoacoustic sensors for CO2 and temperature and humidity obviously has been around for longer than anything else. We're pretty comfortable with their capabilities. I think it's really interesting though, now as we're starting to move into that era of pollutants of specific interest, there's some really good work going on by Ben Jones and Max Sherman and people like that around the main pollutants of concern we see in the built environment, and there's no surprises, really surprises really.

Simon: 53:20

It's particulate matter and formaldehydes and combustion related pollutants like nox and radon and ozone all the stuff we know um, and it seems like we're getting to that point where those sensors have some capability in that environment and use um. So we're starting to see them appear now interestingly on a few of the manufacturers boards. You know some will do ozone, others will do nox, others will do something else, but it's starting to see it. It's interesting. Be interested to see what your community thinks and how that develops over the next few years I think often the community they like.

Achim: 53:56

For example, there are certain years, I think often the community they like. For example, there are certain mox sensors out that, if you like I don't know co, co, no2, um or so on, like four or five different parameters, but when you actually look at at research papers that tested them, they actually call them random noise generators. So yeah, um, they are actually just not accurate at all. And the misperception that we often meet in the community is that they think as soon as they get a value out of a sensor module they just think it's accurate. And that's often a real challenge for us to really explain that a lot of these are the limitations of a lot of these sensor modules very difficult to actually use for non-scientific purposes because you need to do a calibration before you deploy them.

Achim: 55:04

But for calibration you need a reference instruments. Who has a reference instrument at home? And often they also have very fast drift. So even after a couple of months you might need to, you know, like, co-locate them again with a reference and double check if they stay accurate. So I think that's one of the challenges we face, that often the community comes and say, hey, let's do formaldehyde or this or that and we say, yes, theoretically we can get data out of it, but how as a company, can we actually guarantee that this is easy to deploy and still accurate? So that's often the challenge there. The sensor modules we use at the moment, let's say, in the default configuration, is something that's proven. It can do automatic baseline calibrations for CO2. And we can have a reasonable assumption that they work very well and can give you valuable information and can give you valuable information. But I think that's a challenge, often with a couple of sensors that are in the market as well as new ones coming out.

Simon: 56:27

Are we getting an accurate reading? Can we trust the data that we're seeing? It's a fundamental question that has to be asked. But it doesn't necessarily mean a black or white go, no go if we, if we know we can't trust it. But we understand the limitations of that trust.

Simon: 56:50

It may be able to be applied in a different way than we first thought and I think the the kind of the voc community is kind of going through that with tvoc sensors, like we know that sometimes they can be. They can present worse than useless data for trying to figure out what the chemical composition of the air is around you. But it doesn't mean they're completely worthless if they're applied in a way where you understand the changing condition of a space. And I think this is where I always get really fascinated with risk is that we often see air chemistry and air quality through the an academic lens, where we really want rigor and we want to understand exactly what something is. But there's a whole other way of managing the built environment which is on probability of risk and tolerance of risk and perception of risk, and you know VOCs are a good example of that. You can't rely on TVOCs to give you a WHO level of chemicals in a space. Because they're such low levels, there's too much noise, you're not really going to get anything meaningful.

Simon: 58:04

But that's not how our exposure to pollutants occur in a space. It's more to do with severity and frequency of exposure to events. That's the interesting bit. So can we apply that data in a useful way? Very similar to particulate matter? The vast majority of exposure to particulate matter is through frequency and severity of events, rather than a steady state background rate of exposure. So by having these open conversations and being honest about the potential of devices, I think that's really interesting because it opens those honest conversations. Annika, how did you find yourself associated with these guys? And? And because originally you were based out of cambridge, was that?

Anika: 58:50

yes, correct. So I um, I did my phd in cambridge, worked a lot with low-cost sensors there. It was a lot about the testing of these sensors but also deployments. Our biggest project was in China, but we also had smaller projects in the UK and Kenya, germany, and after this I actually returned to Germany. That was COVID times. So I just took the next best job in chemistry.

Anika: 59:22

But I stayed in contact with my colleagues or now dear friends in Cambridge and they told me at some point that they started working with AirGradient and they told me about this open source approach and impact driven, educational driven.

Anika: 59:41

So I got a bit curious about it and, even though I had a job, I asked if I could get in contact and I started talking to Achim and just wanted to hear a bit more about this kind of approach and eventually, actually unintentionally, that led to me working for a gradient and I was really happy. I am really happy with this job In this chemistry job I had. Before I worked in research, but with chemical industries and there was a lot of secretiveness, so basically the opposite of now. So you couldn't even take a group picture in the lab because the people were afraid that we were disclosing some secrets of their industry. So I'm so happy now that I can just openly in this interview, for example, just can openly talk about anything and don't need to be scared that I'm saying the wrong thing. So, yeah, in this aspect, and also, you know, I'm back to my original research again and more closer to environmental sciences again. So I'm really happy, yeah, about this coincidence basically.

Simon: 1:00:56

What's your kind of hopes for the business as a whole? Where does this open source approach go? Do you think from here Are you seeing an increase in the business business of people just buying your sensors and your platforms and wanting to understand air quality? Is there much more potential in the open source end of it? Where do you kind of see it going?

Achim: 1:01:20

So we feel extremely well where we are at the moment. You know, I think I mentioned it earlier. So there was a period within the company where we were kind of finding ourselves, and I think now we are super happy and also confident on what we do. It's also great to do something new in a space where maybe the last couple of years there was not that much innovation. So we definitely have no plans to change. I think what we want to achieve more is that people really see the benefits. You know, I think to some extent there's still a lot of like people just don't know we exist and then they might, you know, maybe buy something and later they find out, okay, they, you know, maybe buy something and later they find out, okay, they could have done much more, like with a more open product. I also do hope that we can tap into the more mainstream market as well and increase the awareness for the benefits of, you know, sharing data openly, for example. So it's to some extent, I think, important that, like what Annika is doing, for example, that the educational aspect needs to go hand in hand with the business side. Educational aspect needs to go hand in hand with the business side. You know that you also like it's really like making an impact. I think that's at the core of what we want to do, and you don't make an impact by just selling a monitor. That's not an impact. Impact by just selling a monitor, that's not an impact.

Achim: 1:03:18

An impact is if somebody changes their behavior and breathes healthier air. That's an impact. I think that's still a big challenge in the whole industry. Let's go beyond collecting data but triggering behavior change, and I think that's very, very complex. It's still something that's very, very hard. It goes into all levels, it goes into communication, it goes into awareness, it goes into understanding air quality and then also understanding what can I do to improve it, you know. So that's ultimately, I think, where, where we want to go and we just feel to maximize the impact, we really need the community, because they are like multiplicators, um of of what we want to achieve and by having this um open approach, we we just get the a huge community behind us, um, and that helps us in like really making that impact yeah, and and structurally, what does air gradient look like?

Simon: 1:04:40

do you do you have regional sales people? Do you have people out there on the ground knocking on doors of business and institutions and like, how's the business, business part of that work for you? Because it can't can't all be academic research and open source stuff, like at some point you've got people, have got to get out there selling and marketing and putting your brand forward, haven't they?

Achim: 1:05:07

interestingly, not so much. So we don't employ any traditional sales person, not a single one. We have a strong science team that Anika leads, which basically has a lot of like a very, very strong network into universities, which is one customer segment. And then we also have like a couple of people we call community managers that basically, for example, write stories about what impact we do on the community and then we put them on the blog and they are reaching out to influencers, to YouTubers, and telling them hey, we are there, do you want to review our monitor? Do you want to talk to us? Basically what we do today as well. And then I think the community is just a sales force by itself, basically Because they see the benefits from the open approach. I think there's also a strong desire from many, many community members to support that kind of business model, so they tell their friends and then they become customers and so on.

Anika: 1:06:32

And then, of course, we have the technical people as well, who help us with the development of the software and the hardware.

Simon: 1:06:37

Yeah, I think what's really interesting about that approach. I can't tell you how many conversations I've had with manufacturers and companies at the other end of the scale the ones making the standard sensors, the ones selling air quality standards and processes into the built environment. They're stuck at the other end of the spectrum. They're stuck in the Deloitte headquarters building in London and the LinkedIn headquarters building in Dublin and the Ernest and Young building in New York. The Ernest and Young building in New York, the glass-towered, headquartered pinnacle of the built environment, that have got Well, platinum and Lead Platinum and can throw endless amounts of money and resources and marketing at these kind of ventures. And that's great because it shows the potential of spaces, how far you can go. But that's not because it shows the potential of spaces, how far you can go. But that's not the built environment.

Simon: 1:07:37

The vast, vast majority of the built environment are offices above shops on high streets and school classrooms and hairdressers in town and little restaurants out in the country and every other facet of the building and homes, obviously.

Simon: 1:07:54

And the challenge those bigger organizations have is getting to those communities because it's a harder sell. There's less profit in it potentially. If you're doing the hard miles in middle America trying to convince school boards to take on air cleaning and air quality monitoring. It's a tough sell. It's a tough job and you wonder if coming at it from the other approach opens up that long tail of the built environment that we're not accessing very well, because the value at that end of the built environment is that is where the low hanging fruit is. It's where we can have the most impact on people's lives and health and well-being, typically in those headquartered buildings. You're just tinkering and optimizing something that's already pretty good of a bank in middle america that the hevax system broke 10 years ago and nobody's fixed. That's where the potential is to make a difference to people and I and I do wonder if this kind of thing opens up that opportunity in some way that others may struggle to get to it's difficult to tell.

Achim: 1:09:10

I think, uh, yeah, I, I don't know yet I don't think anybody does.

Simon: 1:09:14

To be fair, akiba, I think you know we're. This is what I mean. I I think the sector is stuck in a gear at the moment post covid, and it's it. It feels like it's within reach, it feels like the sector is ready to kick on and and progress indoor air quality, but it's just not quite clear where, how, that is yet, how that's defined, um, and I think I I agree with that, simon, and I think that's where the benefit of the open approach comes back in, because it just gives you a lot more options.

Achim: 1:09:50

Basically, yeah, like you're not, you're not locked into a specific data platform. That only gives you a lot more options. Basically, you are not locked into a specific data platform that only gives you this and this and this features. No, you can basically do anything with the data and we make it accessible and this hopefully opens up maybe new use cases which then have a clearer benefit bodybuild environment that maybe are not there yet. So that's maybe one hope that we have as well that by offering an easy, accessible, open system, basically that others can build on top of that and and then maybe develop these um more impactful use cases. You know that that you mentioned um, but I I don't really have a good example yet, but I think that's part of um, our mission, to to enable that kind of um development.

Simon: 1:10:56

Yeah, some lateral thinking and seeing if we can, you know, get several streams of opportunity moving, not just the traditional sas model that seems to be being deployed at the moment, which made me think, actually, are you seeing the same kind of innovation and openness around apis as well, the ability of your hardware and other people's hardware to talk to each other, to cross-pollinate existing systems, crm systems, building management systems? You know, there's. That seems to me to be a huge opportunity as well. There's a, there's a saying you know not another dashboard. Um, that we do get the sense that you know the last thing anybody needs is something else to log into. Um, you know I, we've got enough passwords to remember as it is. That's why everything of mine is on Google God help me is because everything's single sign-on. But I do get the sense that there's huge opportunity in our ability to start talking to other systems and talking to other hardware platforms. Is that something you do as well?

Achim: 1:12:08

or look at yes, so we are completely open. There are a lot of different ways to access the data, so our monitor can actually also, with the default firmware, just send data directly to any MQTT broker. It doesn't go through our platform at all. We can go bankrupt and this monitor would still happily send data to your own server, for example. We have a very, very performant API and some of the endpoints don't even need a token, so you can go on our website, like, for example, for our, like the outdoor monitors that the community publicly shares.

Achim: 1:12:49

There's one endpoint you can just put in your browser and you get a JSON data stream with all the current outdoor data from around the world and you can use it. We have also a couple of customers that are actually building exactly on that. They are developing often their own specialized dashboards, for example. So we have one company that specializes in just report generation for building consultants, for example, so easily accessible and they can ingest it into their own data platform. They can really build the use case around that. And then there are a couple of other companies that do exactly the same. They build their own dashboard for very, very specific use cases.

Simon: 1:13:50

Okay, I'll share a couple of links in the podcast notes of what an api and a token might be, because there's probably a whole bunch of listeners that are very quickly turning off when we start talking about data and that kind of stuff. We have to. We try and keep it as plain english as we can, but, um, I think it's a really important element to this is that how we, how we cross-pollinate these systems, how we generate real value. You know the value may not be in a, a number that's shown on a screen. The number may be in a red flag that's generated on someone's works order program to go and fix something. Then, all of a sudden, that data has genuine value for somebody. They just create a little piece of code that says if these conditions are met, I need to send somebody out to go and fix something you know and all of a sudden that makes it worth it, rather than there being something that somebody has to log in at.

Simon: 1:14:45

Look at a ziggy zaggy line on a screen and see if it's crossed a red threshold somewhere. You know may not have much meaning to most people. You know, I think it's crossed a red threshold somewhere. You know may may not have much meaning to most people you know. I think it's a really interesting area of development that you're interested. You're involved in some really interesting projects as well. One the reason I know you is through my links with henry burridge and the guys doing the sammy project, that's. That's a project based in the uk that uses your sensors, isn't it? For classrooms and air quality in schools.

Achim: 1:15:17

Yeah, that's correct. So that was actually one of the first big projects we got, I think around two and a half years ago, roughly. So SAMI stands for Schools Air Quality Monitoring for Health and Education and it's a huge project. So we deliver 2,000 indoor air quality monitors. So for that project it's the biggest study of classroom air quality anywhere in the world.

Achim: 1:15:48

And I think what makes it really unique is that the researchers try to do two things at the same time, which is also challenging. So on the one hand, it's the scientific part, to understand what's really the air quality around many, many, many schools in the UK. What are areas of concern? Many, many, many schools in the UK. What are areas of concern? What are peak CO2 values that you can see? How does that spread across different building types, for example, and so on?

Achim: 1:16:25

And the other aspect is that the SEMI project actually developed their own dashboard for student engagement. So there's this kind of student not citizen student science aspect to it where they can actually make experiments with the monitor so they can actually put it somewhere and then they can do exercise or jump in the classroom and see how the CO2 rises and then open the windows and see how it falls, and experiment, maybe creating some vocs with different, you know, like um chemicals, or just peeling an orange and seeing how how this thing reacts. So, so, um, and that's this, I think, a super interesting aspect of the project to to really educate um young people on what is air quality and what is good air, what is bad air, how can I improve it. Um, and it's just, the scale is just um astonishing. If you have 2,000 monitors in schools, you can reach a lot of students, and that's also very, very aligned with our mission, and I think that was probably also a natural fit to be part of that project.

Simon: 1:17:49

Yeah, people can go back actually and listen to Henry Burridge on the podcast.

Simon: 1:17:54

I spoke to him, I think, just christmas or around the new year or something, um, and a fascinating insight into both the the technical goals of sammy, but also the citizen science element, elements of it and, and for me, um, air quality captures my imagination in the same way as it does kids.

Simon: 1:18:15

You know it's got a bit of everything. You know air chemistry and physics, and engineering and social science and you know, if we want to excite kids about something that's tangible and that they do every day and that they can have an impact on air quality is great. You know it's interesting enough because it's invisible but can be presented in a way that is really fascinating to kids. And, like you say, the idea of kids running around the classroom generating CO2 and then looking at how it disperses when they go into the other room, they rush into the other classroom and look at the monitor and watch it drop and they're starting to understand concepts around co2 decay and how you might then assess what the air change rate of a space is, and some really great first principles, ideas around air quality it's brilliant. How's that? Are you still deploying sensors for that project, or is that project largely completed from your perspective now?

Achim: 1:19:16

No, from our perspective, we delivered all of the monitors and they are responsive for distribution. So I'm not sure if they got all deployed now or not yet, but they have a really really strong and capable team behind. So it was really interesting for us also to see how quickly they were able to develop their own dashboard and come up with these experiments and really breaking down and I think that's often the challenge that I think also Anika has to deal with how can you actually break down these scientific concepts and make them understandable for, like 11 year olds, you know? And, on the other hand, if you really manage to do this and to learn this, then you have really this avenue um towards the, the general public, and to really um go and and make them understand the importance of air quality and also, like you know, the means to improve it. And so that's why for us, it's also really really interesting to observe these kind of projects. Um, because we are pretty much on the same page or we face the same challenges.

Achim: 1:20:39

You have a VOC index of 350. Okay, so what does that mean now? But on the other hand, it does make sense just to have a number on the dashboard. But on the other hand, it does make sense just to have a number on the dashboard. So you, we have some really interesting ideas and also actually partnerships in place.

Achim: 1:21:21

Maybe one thing to mention is interesting so there's the Energy Policy Institute at the University of Chicago and they came up with this air quality life index and they have a really, really interesting website and you should share the link with your community where you can actually search for, let's say, your hometown, for example, and then it tells you, for exactly your place, how much of your life expectancy you lose because of PM concentration, for example. So if you go on the most polluted places like, for example, india, you actually lose around nine years of your life due to air pollution. And this gives the whole thing a completely different meaning than just saying, hey, you have like 37.5 micrograms per cubic meter. You know nobody understands this, but if I tell you you're actually losing nine years of your life, yeah, okay, then you have an impact. And there are a couple of other, I think, useful tools to support that kind of communication and our plan is to bring that step by step into basically into our dashboard, into our communication with the people that measure the air quality.

Simon: 1:22:46

Yeah, that's a really good point and we did that very successfully with something like smoking. It's's that translation of quite complex epidemiological population level science data into something that translates meaningfully for people that whole disability adjusted life year metric, because it it creates a standard by which you can gauge both at a population level how harmful something is or a risk is, but also at an individual level. We saw that translated with smoking into each cigarette costs you seven seconds of your life, which is just a direct translation of a something you do that has risk, what that means to you actually tangibly, and I think we need to get better at that in the air quality community. Quite how we do.

Simon: 1:23:41

That is not clear yet, but we're getting closer. I think there's some really good straight lines now being drawn between air quality outcomes and health outcomes, certainly at a population level. That make those discussions much clearer, which is great.

Achim: 1:23:59

Maybe adding.

Achim: 1:24:00

One thing is interesting on that co-location project I was really surprised how many universities, also really top-notch research institutions, are just willingly to help us and we don't have to pay them any money, and I think one of the reasons is that they know that all the algorithms that we develop together with them will also be open source.

Achim: 1:24:24

You know so often these algorithms to kind of fine-tune your sensor modules is what what a lot of companies regard as their trade secret, because that's the only difference they can actually make to their competition, because they all use the same sensor module or plant hour. And, by definition, all these algorithms that we develop will be open-sourced or are open-sourced, and that makes it so much easier for research institutions and that's maybe the natural fit that you mentioned before, where research is much more open than many, many traditional companies. But because we have that fit with them by being open as well, that becomes a very, very natural collaboration and then we can actually augment what the research institutions often cannot do, because we can deliver monitors in in big quantities, you know, and they need that for research, and I think that's where the whole thing really comes together and becomes like, like or creates a lot of synergies um that, like then the whole public can actually draw up on and use.

Anika: 1:25:36

Yeah, basically a win-win because we're gone, like the researchers would have to do this calibration work anyway. So basically we do it for them and they provide the references.

Simon: 1:25:47

Yeah and, like you say, I mean most, most electronics industries is just a race to the bottom. You know it's a you're making it faster and cheaper and lighter and quicker and you know it only goes in one direction and I think that's one of the pressure. One of the pressures that's been brought to bear with the traditional multi-sensor manufacturers over the last couple of years is that that they're losing ways to differentiate themselves from each other a little bit because they all use the same three or four major component manufacturers. So eventually you're just buying a white box with a bunch of the same sensors in it and nobody has really cracked the nut of how to translate that into something with a real USP. Everybody's got a dashboard with a heat map and a ziggy-zaggy line and red flags and you know, unless somebody's really logging into it, it loses a lot of its use. And if you can produce a report, great, or an email alert, great. But that only gets you so far and at some point you've got to turn that into from insight into actionable insight.

Simon: 1:26:58

It's one of the interesting things when you look at the social housing sector is that they're probably the more advanced of the sectors in translating data into something that has a value to tenants and landlords. You know, fuel poverty, void risk, condensation, damp and mold, ventilation performance all of these things have a day-to-day impact on improving the lives of people in those homes. But improving the efficiency of the company delivering those services all of a sudden that data has real value then in workplace, because it's such a split incentive type arena. You've got landlords owning the shell and tenants managing the inside of the building. You've got FM companies delivering services. Nobody's really sure how to take that data and turn it into something that is valuable to everybody. Turn it into something that is valuable to everybody, you can, but it's a very long and hard argument and conversation to have with them and I think that's the challenge in that workplace sector at the moment is getting it forward and into places where they know what value it's actually bringing. Beyond getting a well certification or a reset certification or something like that.

Simon: 1:28:14

Day-to-day what does this mean to us as a business? I think that's the interesting jigsaw puzzle they've got to fix. What do you think's next for the sector? What do you think shifts it onto this next year? Maybe I'll ask you first, annika, shifts it onto this next gear? Maybe I'll ask you first, annika. I mean, what do you see is likely to come next that moves this indoor air quality?

Anika: 1:28:41

conversation on Well, I would hope that that's more like also going in the direction of what our goals would be. I hope that this, being open source and open hardware, could kind of also have a role model function, so that we prove that it is possible to be a company who openly publishes everything, shares everything and still survives. I think that is I mean, we're not the only company who does it, but I think that has a nice social factor or aspect to it and I would hope that this might inspire other companies if it's air quality or also other technologies to go down a similar route. And concerning indoor air quality, I would hope that. Well, I guess the sector is moving towards, as you say, smaller, like more integrated in daily life. I can imagine so perhaps at some point we'll all have an air quality monitor in our smartwatch or something similar like that, telling us how to protect ourselves from air pollution.

Anika: 1:30:07

Yeah, in this sense, also a higher awareness of indoor air quality. I've heard a lot of cases now including myself, to be honest of people who were just not aware how, for example, cooking, how much this changes the indoor air quality, just these little things. And like some people were saying, oh, I don't need an air quality monitor, what is it for? And then you know, the children convinced them to have one anyway and then they were like what, that's? Super surprised, how, how bad it is, and started actually changing their behavior. So I do hope that, uh, you know these little impacts, that the awareness raising um integration in the daily life yeah, here, hear to that, yeah.

Anika: 1:30:53

That this kind of is growing and becomes more normal, I guess, looking into the future, yeah, no, I think that's a brilliant point.

Simon: 1:31:03

What about you, hakim? What would your hope be to see this kind of shift on?

Achim: 1:31:14

Yeah, maybe connecting to what what annika said, I think, um, integration of different data sources and automation, I think, is the key. Because what what we see, what already works is, for example, demand controlled ventilation. It just works beautiful, automatically and you don't have to do anything. So what about we have a much denser sensor network in a city and then link it up with traffic lights and regulate the traffic based on the current air quality situation? I think that's not there yet, but this would be a very, very powerful and also impactful use case that's fully optimized potentially.

Achim: 1:31:55

And I think another thing is really linking up more, like health monitoring with air quality monitoring. So, for example, I think there's some really interesting studies like should you do? There's some really interesting studies like should you do exercise in polluted or during polluted times or not? And there's actually often a higher risk of your health to not do the exercise even if it's polluted and to still do it even you breathe in some bad air. But it could be really interesting like, for example, a use case where your smartwatch is linked to an air quality monitor and basically tells you okay, you should do exactly like run for 35 minutes, because that's the the best optimum between air quality or air pollution exposure and your health benefits. You know so, um, and that's an impactful application.

Achim: 1:32:53

So I think a lot of things is really about like we are linking technologies together and ideally, to a big extent, automize um them, you know? Um, that's, I think, where I hope the the industry is going. Um them. That's, I think, where I hope the industry is going. And maybe one more aspect that we didn't cover that much yet is if you look on a global scale, how air pollution is distributed and the number of monitors globally are distributed, we have by far the highest number of monitors in the cleanest places in the world, okay. And if you look at the, let's say, the most polluted places, like in Africa and Asia, there are countries without a single air quality monitor, you know. So I hope that the industry can create more equity between, let's say, the global north and south and trying to distribute this more equally, because I think we know that the first air quality monitor in a country has the biggest impact. So I think there's a lot to do in distributing the, the monitoring networks, more, more evenly globally yeah, yeah.

Simon: 1:34:21

No, I think that's a really good point. I know, I know I would only probably correct annika on one thing. I my hope for you guys, particularly with an open source approach, is that it doesn't just show that it's possible to survive doing things differently, that there's an opportunity to flourish. Akeem and Annika, thanks so much for taking the time to talk to me. We've just swallowed up nearly two hours of conversation there without even blinking, so it's been absolutely brilliant. Thanks a million.

Anika: 1:34:54

Thank you, simon, it was a pleasure.

Achim: 1:34:56

Thank you too, Simon.