#93: An AI Tool to Streamline Biodiversity Net Gain, with Shashin Mishra (BNGAI)

Transcript

Note: This is a digitally created transcript that likely has errors. Ignore the timestamps as these are not correct.



Shashin Mishra (00:12.789)
Yes, I'm Shashin Mishra. I'm the VP of EMEA for AI Dash. We set up the UK office in 2022, January. And since then, we've been building and providing a product around measurement and improvement of biodiversity and achieving biodiversity net gain in the UK.

Ross (00:34.127)
Brilliant, so we're gonna talk more about that in this conversation. This is B &G AI. Can you just tell us what it is, how it works, what sort of problem you're trying to address with that?

Shashin Mishra (00:46.547)
Yes, sure. So the problem we're trying to address is around helping our users improve the biodiversity on the land that they own or manage. There are different drivers because of which our users need to do it. But ultimately, it all comes down to the fact that to be able to fight the climate change, we need to improve biodiversity on the land just as we've been trying to improve the carbon capture. There is a recent realization, widely acknowledged,

that improving biodiversity has a much more far -reaching impact in fighting the climate change because loss of life is what creates a worsening situation for climate change and being able to fight it long term, we have to improve the quality of life for all kinds of species, animals and plants throughout the planet.

Even though our country looks very green, very beautiful when you go out, it is still one of the most biodiverse depleted countries in the world. And that created the urgency which made the government sign up to the global biodiversity framework. And from there came in the requirements that are now known as Environment Act 2021, which require all developments in England to measure the biodiversity before breaking ground on any site of development. And then...

plan to achieve 10 % more biodiversity than the original baseline over period of 30 years once the construction is over. that's the helping our users, helping all those developers, infrastructure companies achieve that biodiversity net gain is at the core of what we are doing. We are also working with a lot of utilities to help them improve biodiversity on their land. And these are as required by their regulators.

But the bigger market that we're addressing is around the need that has come out because of the environment at 2021. Now the way this technology works and we're taking a slightly different approach from a classical approach around land. So instead of relying purely just on human effort to be able to measure and then plan the improvement, we're bringing in technology and building technology to be able to accelerate that work.

Shashin Mishra (03:02.757)
to be able to support the experts who going to work on the site. So the ecologists, planning officers, or planning managers to support them in measuring and then planning the improvement of biodiversity. For that, we bring in a lot of data. We bring in satellite images, we bring in aerial images, we bring in other information like soil data, ordinance survey information. And then we've built

AI models along with software that allows us to be able to quickly identify all the habitats very, very accurately on the ground and then provide that information over to our users so that they can then go and complete the next set of tasks and find out the exact biodiversity as defined in the regulation and in the metrics much faster and with a higher degree of accuracy.

And as they finish that task, they can then go into the planning side of things, complete the planning exercise, and get started for the 30 -year tracking of all the work that they have to do.

Ross (04:09.967)
So why is, I suppose, a fundamental question, why is it important that we have a technological support for this rather than just doing it, I suppose, by hand or, you know, by humans in a sense?

Shashin Mishra (04:24.981)
Yes, excellent question. Two big reasons really, one shortage of ecologists and the shortage of planners. are the two big reasons. To quantify, we're short by about 40 % ecologists in the country if we want to be able to support all the impacted entities. So from developers, whether it is developers or local authorities, they all require ecologists to be able to either achieve or monitor.

the work being done on the...

And being short by 40 % means that there is nowhere near number of people so that they can get the right expertise and deliver the right game. Technology is the only way to get there as quickly as possible and enable the existing talent in the industry to achieve the scale that is needed to be done. And the second is the shortage of planners. Multiple estimates from different reputable sources indicate about this 25 % shortage of planners as well.

which means that the planners already, whether they're working in private sector or local authorities, they have many aspects to, regulatory aspects to take care of as they're building the planning application. Biodiversity measurement and net gain is one more added to the list. And this is not their core competency. So again, having a technology to give them that data driven and bottoms up calculation, give them the, show them the evidence of

what the developer is claiming or what the application is putting forward gives the confidence to move forward with the application and also know that there is detailed measurement behind what they're either submitting or reviewing depending on where they're placed. That's why technology, bringing in technology is really important because otherwise in absence of it, this problem cannot be solved. We cannot be compliant with what the regulation requires us to do.

Shashin Mishra (06:23.153)
And the short -term effect of that is going to be non -compliance, but the long -term effect is going to be that all the plans that are relying on improving and fighting climate change using this as a method of doing it, that methodology is going to fail.

Ross (06:27.104)
you

Ross (06:32.024)
You

Ross (06:38.477)
It's interesting, I mean, I think I've recently, I've been speaking to ecologists and landscape architects and planners about how their experience with BNG so far. And yeah, I think everyone is, it's still so new and everyone is still very much figuring it out.

And even those, you know, those that have a very good grasp of it are obviously then still working with other people, planners, architects, whatever, are, who's very new to them. So it's such a challenge within the industry. And there seems to be a lot of variation with how it's being applied. And also within local authorities, what expertise there is to actually engage with it positively or to ensure this delivered in the proper way. Do you think B &G AI or technology

like this would be part of sort of solving those problems.

Shashin Mishra (07:30.931)
Yeah, technology will be part of solving those problems. Some of the areas where I can see us helping a lot today is accelerating the work so that instead of if it's an ecology consulting company or a developer with their in -house BNG team, in -house ecology team, they are able to not just do one project at a time, they are able to handle a larger number of projects. With the technology, what they're also able to do is bring in

the biodiversity measurements earlier in the cycle. If they don't have the technology enabling their teams to be able to achieve more, they would be restricted to bringing in biodiversity measurements once they've identified the sites, which can at times also create a business risk. They may find sites, they may purchase sites or acquire sites that are not, where achieving BNG may not be easily possible or may not really be possible without spending a lot more money on credits or off sites.

But by bringing that calculation early on, it can become part of the decision process towards identifying the right sites. And that is only possible by expanding how much work their teams can do by enabling their teams to go out and being able to achieve more and where the technology is that enabler. When it comes to ecology services companies, they are also able to serve their customers much more reliably. As you mentioned, B &G is so new that the entire industry is currently struggling with

How do I estimate a biodiversity net gain planning project accurately? There are so many unknowns, unknowns that local authorities have when they're reviewing it, unknowns that developers have. How much information should I submit at each stage? And as they go through the process, the pricing of the project can go haywire, which can make it much more expensive for the customer. But bringing in technology stabilizes that because it offloads a lot of manual work.

and gives them the ability to be able to deliver at a much better price to their customers without making it too expensive for industry to achieve biodiversity net gain. So it's not just making it faster, it's making it easier and cheaper for people to be able to meet the obligations and meet them in the right.

Ross (09:46.764)
And I guess more consistent if you're always using the same tool.

Shashin Mishra (09:48.903)
Absolutely. Technology always does that, brings in lot of consistency in how things are getting done.

Ross (09:56.715)
Can you tell me, you mentioned about developers, ecologists, local authorities using this, can you just tell me a little bit about some of the use cases and how people different, I suppose different professions might bring this into their workflow and what it might mean for them?

Shashin Mishra (10:12.021)
Yes, so let's start with the developers because they are the ones building the projects. So if a developer is trying to do a development in an area, they would be considering few sites to buy from. Now they've got, everybody's got their own recipe on how to identify the site. Biodiversity measurement is one of the new elements there or new ingredients in that recipe. So being able to identify the biodiversity potential of the sites even before they set foot on the ground.

being able to do that remotely very, very quickly, do that comparison, identifying the sites where achieving net gain given the development plan is going to be feasible even before they've sent an ecologist on the ground is one of the very important use cases for them. But once they've identified the site, once they know that this is where they're going to do the build, then sending an ecologist onto that site to complete the planning application, complete the measurements, baseline, and then submitting that.

is where they use the technology. So this is a system where all of that information is going to reside. They can see what was the baseline for every site they've surveyed as identified by the system. And then for the sites where they send the ecologists, they can see the progress for every single project going on until the submission point. And once it's and approved, they can then use it to capture information around the work that they're actually doing on improving the biodiversity over the 30 -year period.

including any follow -on biodiversity measurements. So as a result, it creates an ongoing source of truth for all the biodiversity measurements that are going to get done throughout the 30 -year life cycle. And evidence of all the work that was being done on the site gets captured, including the field notes because of the field app. And that gives a much higher confidence.

along with the database of all the actions done so that at any point of time, if an audit is done, if a local authority asks a question, they can provide all of that information, not just very accurately, but also very, very quickly without going through tons of files. But they rely on ecologists and that's where our second user user database comes in. Ecologists are their experts and they rely on ecologists today for many, even before for.

Shashin Mishra (12:34.005)
identifying any ecological work that is being required on the site, need for any additional surveys. Now biodiversity is added as an additional item on top of that. So ecologists are then going to go on the site and they would want to be able to complete their baseline quickly and accurately and then go from there onto the planning site. So the software allows them to get the initial habitat assessment done completely in an automated way because that's where our AI comes in and then use that baseline to

complete the condition assessment. One of the benefits that ecologists get at this stage is the Natural England has recommended to get a BNG trained ecologist for getting a condition assessment done. That requires the ecologist to be able to be familiar with the requirements, the condition assessment questions, all the habitat classification. But in absence of that training, they are still qualified to be able to answer questions. If an ecologist is asked how many species are in this grassland, they can still

a habitat ecologist can answer that question. So our system bridges that gap for them because it takes away the requirement of being familiar with the latest version of the metric, being from being requiring colleges to carry the question condition assessment questionnaire on paper or an iPad with them. They can just open the app and app asks them the questions and gives them the answers to choose from. So it simplifies the process significantly, allows them to come.

complete the information, it also removes the need for digitization because everything is getting digitized. It's all digital. It's all captured in the system and the calculations are done as they are answering the questions. So they can complete the assessment, come back and they can immediately start with the planning, which is again in a very interactive way. Overall, the biggest benefit that as a result ecologists get is that their effort to complete a BNG project goes down by about 60 to 70 percent, which means they can do three to four time number of projects using it.

But in no way this is trying to replace them because those tasks that the ecologist has to do can only be done by them. It is trying to speed them up, be a collaborative AI and help ecologists achieve more. And then once the application is submitted, then it goes to the local authority. And local authorities are then required to validate the information that has come in, look at the baseline, agree or disagree with the baseline.

Shashin Mishra (15:00.917)
And if they agree with the baseline, look at the plan, review the plan and see whether the plan is correct or not. And there, the metric file itself gives a lot of tools for the local authorities to validate whether the 10 % net gain has been structured correctly or not. What the metric file does not do is tell them whether the baseline on which that 10 % net gain has been structured is correct or not.

So that's where we can come in again. We can help them know why the baseline is accurate, provide all that evidence, and they can make a quick decision on whether the baseline is correct or not. Because unless the baseline is correct, you cannot be sure about the 10 % net gain plan. The plan is not right until the baseline.

Ross (15:46.307)
Yeah, absolutely. No, thanks for that. That's really clear. And thanks for getting ahead of that key question that I think people have, which is like, this trying to replace ecologists? And obviously not, because there's just points in the process where you need trained people to go in and do the work.

Shashin Mishra (16:02.301)
Yeah, we keep hearing about AI trying to take away jobs. I don't think that's going to happen. Not anytime soon, if it does at all. and especially in this case, is the AI can only go as far it can these. It's not really an intelligence. It's tools that there are smart tools that allow us to find out things, provide us the information they go as far as providing an insight. But.

where the human comes in is the expertise. The way a human can apply, the expert can apply the context, machine cannot do that. And that's why we need those ecologists. And that's also why it is able to speed them up because I think as we're trying to solve these increasingly complex problems, in this case, the problem being, how do you achieve biodiversity net gain for the entire country? We have to use experts to do the more thinking job, make, be doing more decisioning job.

rather than walking the field and doing the manual labor. So the colleges in this case can use the technology to let the tech do that walking the field job for them and then bring them the information that they can then use to apply their context and make a decision. If this is a site, what is the best way to achieve a net gain? What is the approach to planning, completing planning application from biodiversity point of view? That's where their expertise then comes in.

So they're spending their time on more higher value tasks and letting the lower value tasks being done by machine. And that's how I think we're going to see this AI evolve across industries, but definitely in this area.

Ross (17:44.483)
that makes sense to me and I think maybe.

would be actually very relieving for some ecologists to sort of let go of some of that time consuming work that probably isn't very, you know, is, will be great to speed up the process. Something I'm interested in is, is the BNG AI, it's obviously does the baseline assessment, does it then provide any intelligence on how you could achieve the net gain, or is that part left to the trained ecologist?

Shashin Mishra (18:14.793)
Now it does provide support for that. So we do the habitat identification first, which is identifying and classifying all the habitats at a very high resolution on the ground. Then the condition assessment requires the ecologist to go on the site and identify the condition by answering those questions. And there is a mobile field app, field survey app to support them during that process. Once they've done it,

And once they've got their development plans on the site to know how much biodiversity is going to get lost as the development happens, they can begin the planning exercise. So it's a very interactive system at that stage as well. They can run the scenarios. They can see the net gain going up or down based on the changes they're proposing. The scenarios, if a scenario they're adding does not bring the effect that they thought it would, they can remove the scenario and add another one.

So for majority of the sites, it's a matter of 20 to 30 minutes to complete the BNG plan, regardless of the complexity. And because now they're interacting with the system and they think, okay, I know that these are the kind of habitats changes I can make. So let me start with that and then start finding other areas. They don't have to run any calculation. They're not entering line by line in a metric file because...

The way metric files designed is for every habitat they're changing, they're supposed to enter a row in the Excel sheet. And there are many, many columns in the Excel sheet that have to be filled in. But here, they're visually selecting a polygon and saying, okay, I want to change this to something else. Or they're drawing a shape and they're saying, okay, I want to create this habitat here. And if that overlaps with five different habitats, which require five different rows to be entered in the Excel sheet, all of that will be generated automatically.

Once they, on the interface, if they say, okay, now I'm happy with my plan, submit my plan and generate the metric Excel sheet for me to send to the authority, it will automatically populate every single row. So that effort is significantly reduced because metric is generated within a minute by the system, but they are creating all of that information in a very visual way by actually looking at the map and interacting with the map. So it does take away a lot of effort. provides a lot of intelligence.

Shashin Mishra (20:37.509)
And I mean, this is still very early stages. I believe very early stages for our product. At this stage, we are able to provide these benefits, but we are now working on other intelligent elements as well that can help them further speed up things like this planning.

Ross (20:55.683)
That's fascinating. think so you could use the tool to support the design process as well. Master planning team could look at different options straight away within BNGAI and it would tell you is this meeting, know, is this on track to meet the net gain or are you you way off?

Shashin Mishra (21:02.25)
Yes?

Shashin Mishra (21:14.837)
Yeah, absolutely. It shows them exactly whether you're meeting 10 % net gain or not. The 10 % net gain itself has to be done across three different categories. So you've got water courses, you have to improve them separately by 10 % if there is any water -based habitat in your site. If there are any point habitats, trees, free standing trees, you have to improve 10 % for them separately. All the linear habitats.

which includes water course as well, but hedgerows, lines of trees, they have to be improved separately as well, plus all the area habitats. So it shows you by category whether you are improving 10 % in each of them or not. So you can say that, okay, I've achieved 10 % net gain on area habitats, but I'm still short on linear habitats. So I need to focus on that. And then you start focusing on that. So it gives you that consistent feedback as you're doing the planning. So you can do the right.

focus on the right elements and complete the plan to meet the requirement. As you're making the changes, it shows you how much time is each change going to do as per the metric. So it runs those calculations in real time. So you know, if you've got a timeline in mind, you know that, okay, whether this is fitting my timeline or not. And if you're making any change, which has a red flag within the regulatory metrics, it's going to flag that up, immediately saying that if you're doing this kind of change, then these are the additional steps you'll have to

So do you really want to do that? Because you may want to keep it simpler and not make those changes. Or maybe you already know that those changes are unavoidable, so you're prepared for that. But it still flags those things up. These are now the additional things that would need to be done. So it's constantly assisting the user at each stage, making sure they're not missing out on any element. And when the system says, now you're ready to send, they actually have everything ready that they need to send.

Ross (23:09.497)
That's brilliant. That's brilliant. Sounds sounds fantastic because I mean, I said this on a podcast yesterday, but it's something that's been on my mind is, that with biodiversity and again, ecologists and designers are having to work together, should be working together very closely on this. part of the problem is that designers aren't ecologists and ecologists aren't designers, but they need, they sort of, need to combine those two expertise and those two skill sets to really do this in an appropriate way. And some of the feedback I've

Shashin Mishra (23:35.327)
Yes.

Shashin Mishra (23:38.676)
Yes.

Ross (23:39.423)
been hearing from the industry is that, you know, it's become a bit designed by spreadsheets on some sites where people are so focused on meeting the metric that actually they're not coming out with the best design outcomes for the people who live there, essentially. So I think having this sort of design aid in the early stages hopefully should be really useful in terms of allowing designers to take a more visual approach to it and allow

Shashin Mishra (23:48.693)
Yeah.

Ross (24:09.048)
I suppose ecologists to work in a more spatial way as well. And that can be a sort of shared language for them maybe.

Shashin Mishra (24:14.925)
That's right. Yeah, I was about to say that this is if you're bringing in two different sets of experts who've not interacted with each other in the past traditionally then you have to bring in a common design language shared language between them so that they can they can do that and that's an indirect benefit of a platform like this because they can both interact they can both look at it and understand without having to learn the business speak or business language of the other other set of experts

Ross (24:46.144)
sense. And think the other thing that occurred to me as you're talking for local authorities is that many local authorities now have local plans with site allocations in them that are quite well quite out of date or at least a few years old and were determined before BNG came into you know into practice and so they don't you know those sites are not necessarily picked with a biodiversity perspective in mind.

which might create issues as they're coming forward and developers are finding actually it's not really viable for us to develop here. And so that's like the ultimate front loading of B &G, isn't it? Is actually going in at the local plan and the site allocation stage to get that intelligence in and ensure that actually these sites are appropriate to develop and that you can achieve net gain and maybe pushing for higher levels of net gain if it can happen on those sites, et cetera.

Shashin Mishra (25:38.941)
Yes. So we're working with one county council. I wouldn't name any customers in this podcast, but we're working with one county council and that's exactly what they're doing. When they first talked to us, they said that they have a local plan across all the councils that fall under that counties area. And they had four projects which were getting close to getting the development started. So they needed everything done on them. But once they saw the progress across those few projects.

They're now using our technology to measure biodiversity and create that initial assessment for every single site that's in the plan. And these are over 50, 60 sites that they've now gone and got initial assessment done by us with a view that as any development moves forward on these sites, they're prepared and then the technology would already keep that data in a single place and the rest of the work can then quickly begin.

but it gives an initial assessment of how many habitats are there, how does those compare? And they were all sites where potential development could have happened. Some of them are potential biodiversity credit sites, but then that right use can be identified, a right value can be identified very early on because of that initial assessment being done. it costs, the initial assessment costs hardly anything, so it's just an easy exercise, low budget exercise to do.

but keeping that valuable data in place that they can rely on as the plan over the next two to three years.

Ross (27:11.353)
Yeah, yeah, it makes total sense. That's great. Really good to hear that people are moving in that direction as well.

I'm interested to hear about more case studies or examples of it in use, but I guess an obvious question that I think maybe some people are thinking is actually how accurate and how specific are the assessments? And, know, it can be really tricky for professionals to tell the difference between different habitat types. You know, is it a neutral grassland, acid grassland? Is it this type of woodland or that type of woodland? So, you know, how granular does it get and sort of how confident are you with

Shashin Mishra (27:30.143)
Mm

Ross (27:47.265)
how accurate it is.

Shashin Mishra (27:49.365)
We use very high resolution optical data to look at the ground with additional data sets as well as I said earlier. So in terms of resolution, we can go as detailed as 12 and a half centimeter for a pixel. So the accuracy in identifying the habitats is as a result very, very high. There are some habitats which are trickier to identify as you mentioned grasslands. Grasslands are

quite tricky and that was a learning for me because I'm not an ecologist by training. But then that's where bringing in those additional data sets becomes valuable. We brought in soil data and then that helped us improve the accuracy. But then at the same time, another benefit is that machine provides is as well because as we are training and labeling those polygons, we are identifying, if we identify an error, that gets flagged up.

which allows us to then bring in more higher value data set, ground truth data set. for example, this ecologically, it may not be the best example, but I've already given the disclaimer. I'm not an ecologist. So if the system flags it as a neutral grassland and the ecologist on the ground says it's not a neutral grassland, it is dry acid grassland. That would get flagged up and then that gets added to the training data set for the next iteration.

Ross (29:00.654)
You

Shashin Mishra (29:17.535)
So we know that the system is also able to then take those optical characteristics combined with the data, additional datasets and say that I made this determination before, but this is the one new determination which is now included in the retraining and that allows it to learn better. And what this does mean that we are creating models that are local to England, but that's how that's the approach for vegetation anyway, because vegetation changes over.

over areas, same vegetation across every few hundreds of kilometers is going to get different. So you have to be very, very accurate. You have to have local models. There is more detail around how local we get, but that's part of the IP. But yes, the general approach is to retrain and then use that retrain model to then deploy for the next user, and the next time, as soon as the retraining is done.

Ross (30:09.388)
Right, I guess that's like a key part of AI is that it learns, doesn't it? And the more projects that go into it, presumably, the better it will get.

Shashin Mishra (30:18.485)
Yes, yeah, but we've not waited. So we've been building this since late 2021, early 22 is when we started deploying it. But we've had that head start. The regulation only went into effect in February this year. But by that time, we'd already been building and automating these models to predict the habitats. So by the time the regulation went live, we were already about 95 % accurate in identifying the...

polygons and the habitats on the ground. We've only been becoming better since then. So now we find that when an ecologist disagrees with the determination of what habitat it is on the ground, we also find that they usually come back and say that this determination is not right, but I can see why your system may have thought it is that. So it's the kind of disagreement that they may have with another ecologist in terms of identifying the habitat on

Ross (31:06.087)
Right. Yeah.

Ross (31:10.723)
Yeah, yeah.

Shashin Mishra (31:14.217)
Condition is another matter though. Condition requires somebody to actually observe much larger area and much more closely and then make an, sometimes there's some subjectivity to it as well, but make that assessment of what the condition should be. So that part is left to the expert.

Ross (31:31.511)
sense. makes sense. Wow, it's fascinating. It is really, really interesting. I think having my sort of experience with GIS and mapping is that things are generally quite coarse and you're dealing with datasets that are often out of date. So I think it's quite exciting to have this sort of package of data that you know is reliable and is incredibly high resolution as well.

Ross (31:56.888)
Can we talk a little bit about case studies examples? How has this been working in practice? And no problem if obviously you can't locations or the people involved, but I think it would be interesting if we get a couple of different examples about, how is it working and what's been the benefits for people?

Shashin Mishra (32:15.689)
All right, let's start from big to small. So we're finding that some of our biggest users, and not necessarily amongst the largest developers in the country, they're larger developers too, but even medium -sized development companies, they're starting to use, realize that the technology is making it very affordable to be able to bring in biodiversity measurements.

Ross (32:18.716)
Okay.

Shashin Mishra (32:44.295)
early on in the decision cycle.

Even though it's early days for biodiversity net gain, the first year is still not over for the regulation. No projects have entered into monitoring yet, stage yet. And even then there are instances where we've seen our users struggling with being able to deliver 10, creating a plan to deliver 10 % net gain on sites that they had already purchased. And that's making them realize that maybe it's a better idea to bring it early on in the decision.

cycle of the site purchase because otherwise we may end up with a site that cannot be developed given the biodiversity net gain regime. That site can still be used for providing offsite credits. So there is still lot of value left, because given that their main business is not creating biodiversity credits, their main business is doing the development, that's not the ideal situation. So we had one user who had purchased a site.

They created a, they did the assessment after the site was purchased. That's when we got involved in the process. We brought in their development plans and then helped them complete the 10 % planning. there was in the end, no way to achieve full 10 % on the site or the land that remained. And they needed to go and buy credits. They did not have an offsite available at this stage. And the cost of the credits was roughly in the range of half a million pounds. And that made the project unviable given the amount of the size of the site.

the number of properties they were able to build on it and the whole commercials did not work out. So the project got shut. And now they're bringing in biodiversity measurements much earlier in the cycle to be able to do those measurements much more accurately and then deliver the

Shashin Mishra (34:34.805)
A different example, a much smaller one, where a user got delayed in submitting their application request was they bought a pub which was lying unused for many years. And they read the regulation and believed that they don't really need to submit a biodiversity application, they'll be exempt. And they thought they'll be exempt because the patch of grass next to the pub was less than 25 meters squared.

What they did not realize was that the area outside the pub, the barren land on which the benches were there, because it was left undisturbed for a long enough time, that had become a non -zero biodiversity habitat as well. It had some grass growing on it, some plants on it. So when they submitted the application, when the authority looked at it, they said, you have to include this area in your grassland.

as a habitat and when you add the two areas up, it is more than 25 meters squared. So you have to get a biodiversity measurement done, you have to do a BNG planning done. In this case, they were able to achieve the 10 % net gain, but it delayed the application process by a months. And then being a smaller business, that delay of few months in the process can have an impact. So just doing a quick assessment early on, spending the 100, 200 pounds on it.

can help them identify if there's any risk or not. If I do, need to proceed further with and spend more money on biodiversity net gain or not, and then be more confident in what they're submitting to the authorities. You asked us, asked a question about accuracy. So just want to mention there that it's not, I said we're confident in the accuracy, but what I forgot to say was that we're not just confident in the accuracy we provide.

We are confident enough in that data that if our users have to go to an appeal, we even support them in that appeal. And that's all baked into the initial cost. There's no extra cost for that. We are confident enough that when we are doing the assessment, that assessment will not be questioned because we'll provide enough detailed data behind it. And if that assessment is questioned, we are happy to represent our users there and back up that data and provide evidence at no extra cost.

Ross (36:39.927)
wow, okay.

Ross (37:01.57)
Has that happened yet? It will be very interesting to see once it does, getting the inspectors getting their head around all of this.

Shashin Mishra (37:04.125)
No, yet, it hasn't happened.

Shashin Mishra (37:13.681)
We're working with, we're trying to build more and more awareness around it. So I think Inspectorate and most of the councils are aware of what AI -DASH does, that they're using AI -DASH themselves or not. Yeah. And the, you're finding that most of the ecology domain is also becoming aware of what AI -DASH does. And many are probably looking very, with a lot of interest to see how it all goes from

Ross (37:42.113)
It was interesting just to pause you there on that small site that you mentioned, because I suppose we can often think of B &G as applying to Greenfield sites, but it does apply to Brownfield's urban sites as well. And there's a misconception that often derelict sites have zero biodiversity on them, but actually if a site is left long enough, it can actually be incredibly biodiverse with this sort of Brownfield mosaic habitat.

Shashin Mishra (37:56.58)
yes.

Ross (38:10.776)
pioneer species moving in and be really important for insects and vertebrates and that kind of thing. that's a whole other very interesting aspect of it, which is sort of urban ecology slightly different to, I suppose, mainstream ecology. So interesting that the software deals with that as well.

Shashin Mishra (38:29.299)
Yes, you're absolutely right. Brownfield sites can be very rich in biodiversity if they've been left alone. If meadows are developing there, they can have a very high biodiversity value, which sometimes can be a surprise because the developer may not have expected that.

Ross (38:43.412)
Yeah, exactly, Fantastic. interesting. Any other points or examples? Case studies?

Shashin Mishra (38:54.697)
I think those are some of examples that stand out.

Shashin Mishra (39:02.221)
There is one more use case that we can quickly touch upon, is around, which is combination of this, which is around creating of, creation of biodiversity credits. So we're finding that some of the biodiversity credit companies are also liking to use the tech now because they realize that they can do a much faster planning. So that there are some examples around that too, where being able to identify the initial habitats can give them an early view of how much potential there is. It's a similar problem, but from a different.

as point of view. And their objective is quite different. They're not seeking to achieve 10 % net gain. They want to maximize the number of credits they can get out of the site. again, being able to see and identify the sites with the maximum credit potential could be a competitive advantage.

Ross (39:52.332)
Yeah, for sure. That makes total sense. I think the other interesting aspect of it is when a site is already quite biodiverse and it basically makes it unviable for development there. from being an ecologically minded person, think that's quite positive aspect of the BNG coming into force is that it's steering. If they can get this intelligence upfront.

it will steer them away essentially from wanting to develop sites that should be maintained as a bulls for their biodiversity and possibly steering them more towards sites that are already degraded or not particularly ecologically significant. that's, think, generally is quite a positive aspect of this as well.

Shashin Mishra (40:37.534)
Yes.

Ross (40:40.023)
Brilliant. think that's been a really fascinating conversation. think I've extracted a lot of information and knowledge out of you. So thank you for that. Is there anything that we didn't get a chance to share that you want to talk about now before we wrap up?

Shashin Mishra (40:56.457)
I think we've covered the topics. Thank you, Ross.

Ross (41:00.131)
Yeah, that was great. Thank you so much, Shashin. Really appreciate it. And I'll put a link to the links and everything. But if you just want to tell us the website or where you would point people towards.

Shashin Mishra (41:10.355)
Yes, so the website's an easy URL to remember. It's www .bng .ai. We offer first project free for every new organization that signs up. So if your listeners haven't signed up yet, it would be a good day to sign up and get their first project on the site.

Ross (41:29.856)
Excellent. Okay, great. Well, the link to that is in the episode description as well. And I hope people check it out and get some value out of it. So we'll wrap things up there. Thank you.

Shashin Mishra (41:39.369)
Thank you, Ross. It was a pleasure talking to you.