 The discussion today and in this series is very much about the digital environment, the concept of the digital environment. So this is both in a sense a methodological and a philosophical approach for how we represent our world in digital form. Supporting science, policymakers, businesses, communities and individuals of course, informing how these complexities impact on our lives and society at large. And we think about an arc of technologies from sensors through to data collection to data processing, the analytics and ultimately visualization as we seek to capture the digital environment. And to address these issues and draw on a wealth examples. I'm absolutely delighted today to welcome the team from the decide project delivering enhanced biodiversity information with adaptive citizen science and intelligent digital engagements. And the team are here representing a broad swath of skills and expertise coming together in this in this to meet the demands of the project. And I think the idea of interdisciplinary research teams is so important in trying to address these complex issues and I'm looking forward to hearing very much about that. And if I may welcome then Michael Pocock hello Michael, who who's kindly agreed to give a short overview of the decide project. And perhaps Michael, you might like to just quickly introduce the rest of the team. And then I will hand over to you to give a short presentation please and then we'll have a look forward to discussion after that. Thank you Michael. Thank you Steve. So yes I'm Michael Pocock I'm from the UK Center for ecology and hydrology. And I'm an ecologist, I've been interested in ecology for my research career. And in the ways in which biodiversity is changing the causes for that and what we can do about it why why it's important to us as society as well. And increasingly thinking about citizen science, and the value that citizen science data can bring, but then the challenges in terms of making sense of those data. And also the real requirement to engage with people and the decide project really came through some of that thinking around that I'll introduce that in just a moment. The, the project was, as he said Steve multidisciplinary project it drew on many people from several different organizations and we've only got a subset of the team here today. But I'll just go around and introduce them or allow them to introduce themselves in person. So first of all, Tom. Good morning, I'm Tom August I work at UK CH with Michael and the computational colleges mentioned how new technologies, new methods can help us improve the quantity and quality of data that we collect from citizen scientists as well as communicate back to them. Interesting information about their observations. And then to Susan. Susan Jarvis also UK CH. I'm a quantitative ecologist and my interest is really in how we can use new analytical techniques new statistical techniques to best analyze the wide variety of biodiversity data that we have. And the final CH person is Simon. Yeah, I'm Simon Roth. I'm a data scientist at CH and I along with colleagues are very interested in how we can make the most out of biodiversity data. And how we can use this science to understand the natural world, but also adventuring recently into into digital twins and that sort of work as well. And then on to chat. Hi everyone, I'm just a trick I'm based at the University of Warwick at the center for it is very methodologies for a background in computer science. But I'm mostly interested in aspects of human culture interaction and data visualization and how people work with data and algorithms and how visualization interaction could be a facilitator between between those those. Increasingly interested in the role of data and visualization and forms of representation in society as an interface for engagement and participation. And last but definitely not least Rachel. Thanks my car I'm Rachel Pateman I work at the stock payment bar and Institute at the University of York. And I'm representing the social science team on the projects that was myself, Alison Dyke Sarah West and Jennifer Rao Williams. And we have a citizen science research group SCI and we're primarily interested in the impacts that arise from citizen science and particularly thinking about that from the perspective of participants in citizen science. Thanks very much everyone. Well Michael, perhaps you have a few few slides to set us set us going to do you. I do indeed. So I'll share my screen. And then the mode so you can see everything. There we go. Yeah, so the, the aim of the decide project and Steve you gave it its long title but it's really about this precision or targeted citizen science, and how we've been coded, co designing adaptive sampling, and to think about ways in which we can optimize biodiversity monitoring and use the biodiversity data in invaluable ways. So citizen science or volunteer and collected biodiversity data is really valuable and it's a crucial part of the environmental monitoring space compliments data from professionals from sensors from remote sensing. And clearly with the numbers of people around on on the planet interested in nature and the increasing use of technology and tools. There is so much potential for citizen science to really continue growing and even even more than it is already. So citizen science those data for use in monitoring and there's a few examples of this on the slides there of data which come from citizen science. It benefits people in terms of their nature connectedness and well being, and it leads on to action as well, particularly at local scales, but supports decision makers in developing action as well. So what I think is there with citizen science data is that there's this massive spatial unevenness in records, and that's based at the real large scales. So some parts of the world, some parts of the UK are much better recorded in terms of citizen science than others. And but also at the local scale, people are more likely to record, for example at the edge of towns and that map show their shows butterfly records. So during the decide project, we spent time talking with a range of different decision makers. And one of the key things that came up that they needed was what they lacked was they lacked comprehensive fine scale maps of natural capital. We can do fairly well at the one kilometer scale, we can do very well at the 10 kilometer scale. We can do very well where decision making is really happening. And there's the finesse of that decision making just just isn't available in a comprehensive way. And for recorders they were saying well they lack tools to guide their recordings so that many of them are happy to make records which are the most useful for for science and therefore for nature. But they didn't know where best to go to make those records. So let in the context that we can use citizen science very well for that national production of trends and for local site based recording but there was there was this space this opportunity in between. So to put it another way, when it comes to citizen science the sorts of message I give is that all records are valuable, but we don't simply need more records more and more and more records but we need more informative records places. where we most need those data so if you are a recorder how would you choose where to record. That's where the decide project came in. And as you can see we assembled this amazing team of wonderful people who brought in a real divergence of skills from ecology and statistics and biostatistics and data science. Through to the social science team at the University of York expertise from Earth observation at JNCC and computer science citizen science visual data science. And this was all set within working closely with practitioners people on the ground, engaging with citizen scientists through the partners at butterfly conservation green space infrastructure for Greater London and the North East York's environmental data center. And so the summary of the project is really this that we've got observers making data and we use data from over 100,000 people who've submitted records over the past 20 years of butterflies and moths. And when they submit those records it becomes digital data 18 million records. Combine those data with satellite data in species distribution models to create these fine scale comprehensive maps of distributions. The key things for this was that this included uncertainty and say that then allowed us to create targets or identify targets where it would be valuable, most valuable to get new data, because those would be the places where we would have the greatest uncertainty and a model outputs. And that means that we can then have this spatial targeting of new record effort, we needed to explore ways of doing that in a way that was appealing that was relevant to people and actually led to the action. So we explored that with and evaluated that. But this cycle of if you think of people as as sensors in this sort of context. And we were able to, in a sense, update the information that we can get get from these people is this, this principle of citizen science adaptive sampling, or it's got that digital twin we're creating the, the model of the environment, but then being able to update it in real time. So this is through a few different ways so just running through three outputs one is the decide tool. So this was targeted towards those 100,000 or so butterfly and moth recorders. So they could click on any particular location. This was all developed through a process of co design. And when people see the maps they the colors identify the recording priorities, you can click on to get pins for suggestions based on accessibility. So we included data on accessibility within this tool. This was updated in real time from records submitted by the I record platform. And you could click on those pins or any place on the map actually to get more information about why the records would be valuable from that particular location. So you can see that it was used by a large number of people, nearly a couple of thousand people visiting the tool several times each over the course of the last summer. And importantly, people spent nearly four minutes on the tool with every interaction or that was the average at least. And so clearly people were interacting and using the tool and as I said we're going through a process of evaluation now. But just there's a quote there being able to see how impactful my recording can be directing me to where recording the greatest impact is immensely powerful and motivating. One of the other things that we wanted to explore was that was a tool which people could go to news when they wanted but can we actually proactively push feedback to people and do that in a personalized way. And so 850 people signed up to receive weekly data stories during last summer about the impact of their recording. So these were tools or these were emails which were sent personalized, we actually developed a range of different data stories as Simon and Tom called them, testing the appeal to different types of motivations and our people more motivated by collaboration or competitiveness. And so we're exploring that data now, in terms of the evaluation. One of the one of the challenges with this is that people can go out and they can record butterflies quite easily, but moths. Just that bit more tricky they fly at night and light traps are a great tool for being able to record moths, but typically they either need to be plugged into the mains or they run off generators. So what we did was working with some collaborators over in the Netherlands we developed these portable moth traps, and deployed them in networks, particularly in Yorkshire and in London. So networks of volunteers were going out with these portable moth traps which were, which were running off mobile phone chargers. Power packs mobile phone power packs, which suddenly opens out this opportunity for more moth recording. And then, as I said we, we really focus this around the needs of the data users and so we've developed this tool, and based on discussions with the data users and decision makers, and different scales. We provide this information back to people in terms of these fine scale maps of, of the distribution, and the biodiversity importance of areas. We did this building on different case, different case studies. So thinking about how this needs, or how this would be valuable in being implemented within people's working practices. Finally, Steve, you introduced the whole thing of data and the digital environments that we're in. Well, how, how actually do we make sense of this it feels like quite can feel like quite as cold, quite hard clinical concept. What we want to do is work with artists and so we ran this installation last December, working with a painter and sculptor, and a poet. And so we created this data sets dream. So, and we invited people to join artists and scientists in the woods or liens house gallery to explore what happens when we turn nature into data. So thinking about that relationship of nature in this highly digitized world. If the data could talk, what would it say to us in this time of declining populations of our beautiful butterflies and moths. And this was a really revealing project to run alongside decide, which really I think gave warmth to this concept of, of data and a digital environment. So we'll leave it there and let you Steve. Fantastic. I think people will be interested to know how to find out more about the data sets dream that's fascinating actually idea of bringing artists and poets alongside scientists for interesting. I mean, I'm really struck by your comment about, we don't just want to collect more data it seems we have so much data these days and it's really about trying to decide where to where to actually get those records from so of course your project is very, very wisely named of course. And I suppose you know the citizen science is something that is, is a concept that is becoming aware and that's obviously one or two flagship things like RSP bees garden watch that seem to make their way onto the, the news and everyone's aware that they can collect data but perhaps not always quite understanding what is the challenge for or how, how the data meets the various environmental challenges that that that exists and maybe we could start off by just asking you the team what what the, what is the challenge that you're actually trying to address with with the decide project. And how does it, you know, how does this challenge affect affect all of us our lives and society at large. So if I go on that there's a, I see sort of two things one you just sort of touched on there about a citizen scientist doing the big garden birdwatch and we kind of owe it to citizen scientists who are giving up their time to collect data to ensure they're collecting the data that is most valuable for the questions we're trying to answer. Additionally, it's it's data is close in quite an ad hoc fashion people go kind of wherever they want wherever they want and kind of record what they see. But we know that a lot of people are motivated by their shared ambitions for conservation ambitions that we share as well. And so things like the decide project kind of help to ensure that they really are contributing in in the best way they can. Another thing I'd point to is, you know, the current government has big ambitions for land in the UK has ambitions for building homes has ambitions for conserving conserving land has ambitions for growing woodland area, etc. All these things require decisions to be made about where things are going to happen. And it's, you know, when you try and make those decisions at a national scale, you know, where are we going to pull these things. The empirical data on where biodiversity is can be lacking at the fine resolution, but at national scale, but models can be used to fill in some of those gaps. And really what decided to try and do is try and make those models as good as they can be by getting the data where they need where it's needed so we can make predictions about what is what are in those locations. And you've certainly had a very encouraging uptake and you're Michael you're showing that thousands of people have used the tools that you've put together which is really tremendous. And of course the these tools utilize digital approaches and of course this whole program is about the digital environment the representation of the world around us in digital ways now. And it's fascinating talking to all of the the project teams for our demonstrated projects about the perspectives that they take on digital environment and what that, what that concept means I'm just interested really for for yourselves and for the for the decide project team. How does, how does digital environment relate to to I mean how did you think about using digital techniques to do what you've done in the project. Coming on that. And so I think. I think this project is really interesting because at the core of it is something that's actually not very digital it's people experiencing nature and being out in wildlife and that it's in a way seems a very non digital thing to do. And you know to be outside in the wilderness. And I think what's interesting is that this project has demonstrated that digital technologies can actually add value to that process in quite a lot of different ways so partly that's about how people are recording that data where they're going to record the data that's obviously key bit of the project, but then also how we interact with people around that data collection and also how we analyze the data. So something that shouldn't necessarily be underestimated is the actual amount of modeling and technical data analysis it sits underneath this project that sits underneath the tool that Michael demonstrated. And that's a huge amount of work that really relied on digital technologies high performance computing to be able to do that because it's just a huge amount of data crunching and processing underneath that. Yeah, right. No, I can see you have a whole array of digital tools that you bring to bear on on this. And so really I don't think you know that without these digital tools you clearly couldn't couldn't do the project that's the thrust of it. But I mean there's you've used you've used a number of tools to to do this. I'm just, I'm just thinking whether there's any other digital tools that you would like to bring in and what what what are the opportunities do you think for extending the sort of work you've you've you've shown us. Anyone had a view on that. Yeah Simon I didn't know whether you wanted to talk a little bit more about some of the advances that we're thinking about using AI within these digital models and actually that's sort of the scaling the high throughput the the supercomputer power that's required for these questions. So I think I think what really the really strong district district component of this project is kind of as a sign of how like the state we're getting to in terms of being able to, you know, you hear about 10 years ago said we're going to run 100 species across the entire UK, every 100 meter, 100 meter cell that's that might seem like quite a lot back then but you know we're getting to the point where these the tools that we've got that we've shown in this project. We can run things but then I think it also just shows the opportunities going ahead in terms of new modeling techniques. And there are some things with with with data in the current models. So we're not making use of the fact that the sensor to satellites are repeatedly go around the same piece of covering the same areas of the planet like multiple times. And so you could really, you could really push the real time element of these sorts of tools and I think that's where we're moving towards in terms of. Of course, we're within the sort of the new sort of digital twin agenda, almost that we're getting to the point where we've got social data models we've got lots of data coming in, we've got the computing power to process it and then we're building and we've got the experience in building the digital engagements that people can interact with these and make useful decisions about how they can conserve land or where they should go and record. There's a lot lots of development opportunities and the future perhaps as well. I was also struck, Tom I think in your, in your talk you, your introduction there you mentioned the, the role of the government thinking about competing multiple periods for land. And I think that I think, you know what the role of tools like this and helping to inform policy makers and decision makers is is is key and I just I just wondered whether you're what your, you know, team what your observations are on using tools to support policy makers trying to make actually quite difficult decisions about allocations and strategic decisions about planning resource planning and so on. What's the role of the tools like this and supporting and supporting that. Yeah, I'll make a start and others might want to chip in so we did have some stakeholder engagement with the sorts of people who would use this data so it's kind of local local typically making decisions about where infrastructure are or making planning decisions. One of the things that we're really keen on was was not only providing data on where species are but also that uncertainty uncertainty actually underpins the vast majority of what we're trying to minimize uncertainty through data collection. There's also about effectively communicating that uncertainty to these end users, you know, to say, you know, we maybe we can make some good recommendations about what biodiversity is in locations x and y but over here in some of them were quite uncertain and then that puts the onus on then to contract surveys and those sorts of things, but I wonder if maybe Rachel you might want to say something about these sort of stakeholder groups. Yeah, so my colleague Alison Dyke led the strand of work on engaging with different stakeholders in terms of understanding their needs around what types of data they need. What do they use them for what scale do they need it at but also these challenges as Tom said around interpreting model data versus raw data and that's something that was really highlighted in those interviews and subsequent workshops that we had was. Yes, we can provide people with these outputs but actually what we also need to provide them with is the support in terms of understanding these different types of data that they might not be used to working with and that's kind of our responsibility. So if they need things out there to be able to give them that kind of context as well in terms of how they use that data. I think there's also a point around. I don't we don't at this point expect models to be replacing data, we expect to be supplementing and that to source information that you use alongside each other and they've got their strengths and weaknesses. And what actually the key use cases were identified was the idea of being able to use a model data to then scope where you would then go and collect actual data from from from in person surveys so I think they were the sort of model data in room and collect data in tandem. Okay, thank thanks and while I was actually my next term, I'm intrigued by this idea of sampling in different areas, some filling in areas but I see actually there's a question from David Hopper thank you very much for Hooper sorry and thank you very much for your question and let me put this to you all. So David asks, I submit lots of insect data to I record covering a range of orders not not just Lepidoptera. And I typically target locations that are either close to where they live and work or whether expect to see a wide range of insects and so the question is, how would you persuade an enthusiast like David to go elsewhere based on a subset of interest so you know if you've identified an area that needs to be filled in, how are you going to get enthusiasts like David to go and collect data in those other other areas and what can you do with your digital tools to encourage that. Well I think you might be coming on to talk a bit more about this in terms of the code design elements of this project later in the conversation Steve but I think it really shows the importance of code design and within this project that we did end up with quite a deep understanding of those sorts of things which might motivate recorders Rachel would you like to say a bit more about that. So the strand of work that we were doing at York started with a series of interviews with different types of recorders so these were recorders that we selected through the kind of local recording network. And we chose them to represent lots of different types of recorders so people that like to record over a large area, or just in a local patch people are recording for a long time or a very new to recording. We had an urban center in London and we had a rural center in Yorkshire as well. We had a series of interviews with them asking them about things like how they got into doing biological recording, where they record, how they make choices about where they record and when. And I think what that highlighted to us was there are a huge number of different types of recorder. I think we did 35 interviews and I would say every single person, you could class as a different type. So it's reflected in their motivations but also in their behaviors so some people are motivated because they want to stop a development on a particular patch and they're really motivated by just getting lots of data from that. Some people just want to go out and have a nice time and photographs and butterflies and so they go to places where they're more likely to see nice butterflies. So I think what that highlighted to us was we need to tap into the right type of recorder and this tool that we're promoting which is helping people to direct their recording to places where it's most beneficial for this modeling that's not going to suit everybody and that's fine. Because actually what some of the analysis that we've done has shown is that we only need to redirect actually quite a small fraction of our recording to have quite a big impact for the quality of our models. So if we can tap into that subset of recorders and we can persuade them just to redirect a small amount of their recording effort actually the benefit in terms of the modeling will be quite significant. Yes, I was very struck by the interesting I was very struck by the comments, like I think you made at the beginning about whether you have a competitive or a collaborative approach some people respond very well to a leaderboard. I expect and I think of some other citizen approaches like geograph, where you have to have as many, you have to photograph as many five kilometers squares or whatever it is as possible and the one that's got the most photographs wins you know the leaderboard but other people will focus much more on why they walk their dog or whatever. Sorry, maybe on that point I think quite an interesting discussion that also informed a lot about how we approach the design of this digital engagement that we that we developed during the site. There's different motivations that Rachel talked about and, and, and different ways people get excited about going to a new place doing a different kind of recording is quite different. And that's something we try to explore within the, the might decide part of the project with these individual personalized stories that we send me, but we tried to understand that what kind of narratives would appeal most to different audiences. And that there is a lot of difference and that's, that's so important to understand those differences and try to reflect that in the design I think that there's something that's quite fundamental in the in David's question. How do you encourage that how do you bring what excites people out in those interfaces and how do you communicate the right messages to the right people and that's only possible through the co design participatory design methods that we tried to deploy during the project. I was going to say you're very interested I know in that in that sort of interface between people and technology. And I just wondering what what you what you've, what what you've learned about how you develop that technology to to encourage uptake and use and an interest and enthusiasm about it so how do you, how do you do the magic. It's been such such a fascinating journey and such a learning experience for me I mean when we when we started out we had a few ideas how we might approach that we tried to sort of set up some principles for us, I mean we had this a beautiful a perspective which was that we wanted to display information that's relevant accessible action with an appealing, but we didn't know what those meant and that that came through the research and and what's relevant what's accessible. What kind of information people need is very different to what we initially anticipated. When we started we thought we should talk more about the probabilities and uncertainties and communicate that. But we quickly understood that that's not always what what matters to people and and true dialogue and regular involvement of those who do the recording. We learned how we frame the messages. It's how do we build the right kinds of visualizations and how do we write the right words and how do we annotate these visualizations in a way that people would understand and engage with. I think we'll come back and ask you a bit later what the, what the key sort of learning advice you have for other people starting projects like this but I'm, I'm also struck Michael you mentioned a little while ago. There's not this notion of co design and I think I'd like to come back and scratch away at that one a bit more what, what, what do you mean by co design and how how is that informed your your your design approaches. So chat I began to begin to address that particular thing and I gave a talk at a conference a few months ago talking about the decide project and I was saying that co designers been so important within this project. I mean, maybe the old way I would have done things is to think within a project team. Many of us are interested in recording wildlife ourselves so that's fine. We know roughly the sorts of things we're aiming for. We would design a tool, and we would perfect the tool we would then put it out there for feedback, and I'll get really upset if people didn't absolutely love it. Instead, we adopted this process of co design right from the beginning and developing this and rich back my who's doing a lot of the technical design of the tool said that it was quite a different way for him to work, and he had to adapt his working style to be prepared to come up with very rough wireframes and prototypes and chat I and Greg McInerney from warwick also led us through this process. We're doing it in a very sort of basic way, and then doing it with a trusted group of people. And, and, and so one of the questions I was asked at this conference was picked somebody said well, how, how different would it have been if you didn't do co design. The answer was I genuinely don't know. All I know is that it would have been very different because it's almost like every stage within the process we, we were shaped by, by the, by the, by the importance of the co design by the feedback by the interaction with people. Sometimes that was a couple of people doing a walkthrough interview as they tried to use a wireframe for the tool. Sometimes it was a focus group, which were really valuable for getting different perspectives. And it was also providing survey feedback when we put out the first iteration of the tool, and all of these fed together and just really shaped where we're going so for me it was, it was such a valuable experience to do this. And one other thing I'll say actually what, so one area where the co design did really shape what we came out with was, we thought okay, obviously, and David alluded to it to in this question. We must provide people with as much information as possible. We must tell them as Tom said we must communicate the uncertainty we must tell them why they're there we must think of all the reasons why they should go to this site and not that site and all everything else. Fairly quickly people just like, look, actually, I can make my own decisions. You can give me an invitation. I can choose whether to accept it or not I can I can figure out where the car parks are. Thank you very much. And you don't have to tell me, although we have got that layer of accessibility which is, which is really valuable. And basically, as a recorder, this isn't going to be true for every single recorder but a lot of recorders were saying, we trust you. Tell us where to go that becomes an invitation that we can accept or not. And so, so that question of people trusting was really valuable in terms of shaping where we went. And I think, I think just sort of following that up the, this notion of the citizen as a sensor as an interesting one because that that sounds quite mechanical and impersonal but, of course, you know this this idea of trust and respect going back the other way it has to be absolutely systemic about the design of the tools and approaches you have I'm sure. I'm, I mean, Tom, you've, you've discussed this, this idea of the citizen as a sensor I'm just wondering, how do you, how do you, and others how do you make that sort of interface with the technology that perhaps can adapt to different people's motivations almost in, in, in helping to collect this this crucial data. I have to say, I'm not a big fan of that that phrase is a sense that it's very kind of, yeah, sort of authoritarian and big brother isn't it. And I think that's kind of, that's what we've been talking about, we've been talking about motivation and co design is that this is very much a two way interaction between scientists and citizen scientists, professional scientists. And so, you know, when we're when we're developing these tools, it needs to be, it needs to be beneficial to the user, right. I think, long gone other days where we develop tools which are just like vacuum cleaners trying to suck up data from unwitting members of the public, you know, now we're, we're looking to design tools that are really helpful beneficial to the user. And so, you know, we found through the interviews, etc. that you know people have these shared ambitions for that can conserving biodiversity, they genuinely, you know, the people who are contributing large amounts about a record they want their records be as valuable as possible. And we're creating a tool which can help them to identify where those most valuable places are. I think it can maybe look a bit beyond the side as well to other projects that are kind of incorporated inside to see much more use of things like AI, so tools to identify, so particularly sort of computer vision and acoustics ai's that can help you identify birdsong or can help you identify a butterfly from an image. And these are again sort of technological disenvironment tools that are helping people to get towards the correct answer now they're not replacing taxonomists they're not replacing the human brain. But there are another tool just as a field guide as a tool to help people get towards that identification. There's a whole range of these sorts of apps coming out now and I'm sure they're very, very popular. I'm, if I may, I'm just going to change gear from my because I see Joe Zong thank you Joe for your question and the audience and if anyone else listening has a question please do just pop it into the Q&A. So Joe's asking that whether, whether we actually have the plan already or, or, sorry, do we, do we have the plan or already have the database for recording trees and vegetation so I think, if I understand what Joe's asking you know, can you use these approaches for recording other other forms of environmental phenomena and you know what could you say about the database that's for doing that. Perhaps as a question for some of the data folk here. The answer is yes there are recording schemes across the UK for many, many different taxa including plants, flowers, trees. There are schemes which are interested in the phenology. So that might be the first time you see a flower of a certain species in the spring and that's really important for thinking about the environmental changes of climate change. There are pollution changes of plants of animals. So yeah, and there's a phenomenal amount of stuff, and which is which is going on. There's tools like I record which we host at UK CH, which, which are really valuable for collecting any of this biodiversity data. And Christine sums thank you very much for your question as well on the Q&A. Decide team have you got any insights or lessons to share on successfully engaging volunteer effort, and in particular reflecting on how to successfully deliver co design. Is it all about aligning motivations. What's the secret. Who'd like to have a go at that. And I think many people actually could give different tips on that Rachel would you like to reflect. I think it's coming back to the point that so it's recognising that there are lots of different types of people who you could potentially be interacting with. You're not going to necessarily suit all of those people so it's engaging a range of people in your co design process in the first place I think so you've got that range of perspectives. And then it's tapping into those ones where you feel like there's the most potential for them to be responsive to your tool and engage with them. And I think in terms of lessons learned about the co design process as a whole, which I think has been alluded to already is that it is a different way of working, and it is pretty resource intensive. It requires, it requires time and actually we were quite fortunate because I guess a positive from COVID is that we'd originally planned to do all of our kind of co design workshops in person and we had to to change that to do it all online but actually that turned out to be quite a benefit to us it allowed us to be quite agile and allowed us to do a lot more interactions than we would have done had we been trying to organise in person workshops and allowed us to have people from London and Yorkshire in the same meeting of mixing together. So it allowed us yet to have these kinds of very quick succession iterations of interaction with the recorders, then an update to the tool and another interaction and update to the tool, which was really good but it did require a lot of effort and resourcing and being agile and that can sometimes be quite frustrating because you think I just want to produce this thing and it be out there and actually you kind of have to take the time to go through this. And yeah, so I think it requires buying from everybody in the team and a good group of co designers to embrace that process and go through that process and like Michael say we don't know what the alternative would have been we can't do a test of how effective is the un-codesigned tool versus the co design tool, but I think our evaluation that we've done and the reflections that we've got from the recorders and the project team show those points where it's definitely benefited the final product. But as a demonstrator project, I would think it's clear that the co-designed approach seems to be the one that you'll be recommending for future years as well. Charlie and Simon, one thing I'm curious about you've mentioned digital twins, something I'm interested in and I'm just wondering if we could just come back to this notion of digital twins and what does that concept mean to you in the context of decide and what are the opportunities that a digital twinning approach gives you for advancing the science and understanding here? Yeah, I'll speak to that. So I think in terms of this, the digital twin aspect is the synchrony of the digital and the physical and being able to really create that link is important. And this adaptive sampling idea where the model knows some stuff about where species are, but it's uncertain about other places. And so then through the engagements directly to those places, and you create that feedback loop that I think is one of the key aspects of moving towards digital twins is the interaction between those two. Otherwise they're not twins, they're just a digital thing and a physical thing. So those feedback loops are important. I'm just wondering how would those develop in a project like decide those sort of loops back? Is this sort of helping the clatters to understand the big picture of what they're doing? How do these loops back work? I mean, in terms of looping back with the participant, through sending out emails, I think that's quite an important way of helping people interact with the digital twin. If we're working towards a different approach, then different users will interact with different ways. So say someone who wants to make a decision about where to build houses will interact very differently to someone who's looking to know how to record butterflies. And, yeah, facilitating those links with interactive, like digital engagements, like the Mitre side emails that we tried, I think are really important for that. Chelsea, do you have some thoughts on that? An interesting term I was thinking that really maps here is this notion of dialogue. And I think the way of engaging and going back to people about the value of what they've done or what they could have done differently is the starting of a dialogue between those maybe the worlds that might be a bit apart. And we've heard the mention of trust and engagement. And I think this dialogue is where those values are built up. And for us, methodologically co-design was to build trust and creating an inclusive and equitable sort of place for people to contribute to the project. But I think these digital engagements that we design, I think that they're quite a valuable tool to build that dialogue between, let's say, science and society and giving people informative, accessible information, but also opening up new opportunities for society could be part of science and could contribute to science and then communicating how much value there is in those efforts in participation. You can sort of draw on a familiarity that people have with digital in the rest of their lives to help with this environmental project and there's a familiarity that's the natural familiarity people use their phones. They can understand the technology and you can actually harness that in this project. And not just this project. I mean, Susan, if I may turn to you as well, I mean I'm intrigued and the question we had just now from Joe about other types of recording with trees and vegetation. I'm just wondering what your thoughts are on the learnings that we've had in this project and what the opportunities are for extending the scope or do's and don'ts really for the opportunities for other types of environmental phenomena that we might want to capture and what your thoughts are on that. Yeah, so I think it's interesting when we talk about other fundamentalist I guess we've got two elements of the project really we've got the sort of engagement with the citizen science and absolutely we can do that in other projects with other groups of citizens who are recording other things. I think the other sort of element of the project that is quite transferable is thinking about the methods and the the analysis that actually sits behind this layer of where we tell people they might want to go. So this idea of actually doing the, the modeling and the estimation of that uncertainty to inform our next batch of sampling that's definitely transferable it's something that is being used across other in environmental science domains, something that's quite increasingly common. I think this is the first project I've known of where we're doing this is in science which is quite exciting. But the concepts are definitely transferable I think in terms of biodiversity monitoring and we're going back to Joe's question a bit I think what could be really exciting in the future is thinking about how do we use the broad range of data we have in the biodiversity data landscape so not just thinking about citizen science data but thinking about other types of data that are coming in, maybe that's through new technologies, such as the sort of automated image recognition idea or maybe it's through established traditional monitoring techniques through professional monitoring, but bringing all of those data sets together and then thinking about how we use all of them to inform where to get those new samples where to get the most out of our information I think that's quite an exciting thing to think about looking forward. Great. Thanks. I see we're unfortunately creeping towards the end of the hour so I've got a couple of quick fire questions if I may so brief brief answers on Tom maybe I could start with you you've mentioned the role of AI as well but in this concept of the digital environment. Clearly what what you've achieved you've outlined but what do you what do you see as the sort of the challenges that still that are still outstanding that still remain in in engaging projects like this with digital approaches. That doesn't invite a short answer Steve. To summarize, I think there's a lot of new methods being developed specifically in like AI space. And so a real challenge there is knowledge knowledge amongst our sort of community of those tools that are available. There's a lot of interdisciplinary that interdisciplinarity needed. But also there's a requirement for individuals who themselves are transdisciplinary. So ecologists who understand computer vision and AI. I think there can also be barriers to participation, whether that's access to data or access to compute resources required to do things or access to knowledge. So I think there's, there's a role as well for our community to try and open up knowledge more and make things more inclusive on that side. I think the times come to ask the Rachel if I may and I'll finish Michael with yourself but Rachel, I'm just wondering, as the project progressed. As a demonstrator we've mentioned this idea of learnings for other colleagues in environmental sciences and indeed other disciplines who are interested in taking participating in projects like this in the future. So I'm just wondering if you, if you could sort of sit and give us a view of what some of the best practices are that you've, that you've established or that you've uncovered in your work. I think engaging with the right stakeholders. So as Michael mentioned at the beginning, we have butterfly conservation on board who are really trusted organization from the recorders that we were working with and they facilitated a lot of our engagement with the recording community. So identifying the right stakeholders to be able to access your end user I think is a really important thing. And as I was alluding to time for co-design is really important and having everyone brought into that process, being very respectful of the people that you're engaging with for the time that they're giving up and contributing something that they're probably not as interested in as we are as a project team. And so having somebody dedicated to that kind of communication with those, with those end users was really important as well. And I would say communication again. So we've just kind of started well in the process of analyzing the evaluation feedback from the final survey that we did. And I would say what that's highlighted is that those people who are giving is not so positive feedback. It's probably actually because they've not quite understood what we were trying to do. So despite having been through this co-design process and doing our best to get the messaging right. I think there was probably more that we could have done to invest in actually nailing that. And I think so having somebody on board who's a real communications specialist and who can help with that is really important as well. That's fantastic. Great advice for anyone contemplating a project like this. Well, I mean, thank you all very much. And I mean, Michael, you spoke at the beginning about the range of skills in the team. I think the, you know, I might add clearly that having an interdisciplinary team is another best practice as well. But actually the skills needed to bring everyone together to focus on areas. I mean, Charlie, for example, I understand you're not an environmental scientist, of course. You're, you know, working, working alongside all the other colleagues on how have you, how have you done that, Michael, just to finish a last, last question, how have you, how have you done that? I think one of the things that has been really valuable is not to have things to siloed within the project as a research project. It's quite easy to go, okay, the computer scientists, stop you go and do your thing. And social scientists go and do your thing and then we'll come together at the end. But we've had such regular communication. I think the team has been large enough to really have that breadth, but small enough to enable a huge amount of interaction through our regular meetings, but also through sort of sub task meetings, where we were genuinely trying to get this, get this engagement across disciplines. And I think bringing in people who have that expertise and engagement as Rachel was saying, and not undervaluing that. So we ended up with someone part time who did engagement on the project and they were absolutely superb in terms of being able to communicate in the blogs and the Twitter and engaging and once again bringing people so much together and working with the partners. But all of this I guess is also a lot of it is driven by the fact that there's a huge number of people out there who really care about nature. Who want to do something. And the citizen science is a way for them to engage with nature. And then you've also got organizations and governments at different scales from local to regional, who are faced with this, as Tom said all these competing pressures for food change for biodiversity for health for well being all these sorts of things and being able to bring these together and and have a small step towards supporting all of these agendas I think it's been really motivating for us as a team. Well, thank you for that. Almost sort of a closing closing statement there. I'm afraid that's really all we've got time for this week and looking at the clock. And it remains for me to thank very much all of all of you the panelists who've joined us from the decide project thank you so much for your time thank you very much to the audience as well and for the questions that came in that that's great. It's been a great discussion as we've we've gone through trying to understand how digital environment is used in in this context for citizens science and being some great great examples of that so thank you very much.