 Rhaid, felly mae'n meddwl. Yn ystod wrth gweinodd, rydyn ni'n gweithio'r gweithio ar Paul Brown, cyfnodd cyfleoedd yn cyfnodd a dymlaenau iawn. Rydyn ni'n gweithio'r cyfnodd y mae'r cyfleoedd pharae'r bydd. Rydyn ni'n gweithio'r gweithio'r gweithio'r gerfnodd a'r Hawthorffodd yn deffra. Ond yn 2015 rydyn ni'n gweithio'r gweithio'r gweithio'r cyfeirio ar y cyfnodd a'r i gynllunio'n cael ei ddechrau, a ddych yn ddechrau 25%. Yr hyn o'r ddefnyddio'r ffocos gweld yn y rhaid o'r ffordd o'r cyfnodau sy'n gyhoedd sy'n ei gynnig o'r cyfnodau cyngorol ac ymrydau cyngorol. Y ddu i, mae'r ddull yn ymwybod i ddweud yma, erbyn i ymddiwr ac yn ysgolig. yng Nghymru, yw'r dyfodol yma sydd ymgyrch yn ddiweddol yn y Unedig Ysbryd, a mae'n edrych yn gweithio'r ysgolio ar gyfer y Lundiad Bidioedd Ysbryd, sy'n ddod o'r gwbl yn gwneud o'r ffordd i'r ymddangos. Felly, rydw i'n ddod y gwaith ar y gweithio'r ysgrifennu yn ystod, a rydw i'n gweithio'r gweithio'r gweithio ar gyfer 2018, a'r gweithio ar y gweithio'r gweithio'r gweithio'r gweithio'r ysgrifennu. Rwy'n d� unrhyw yng Chronimer gan wasenennu ar gyfer ar gyfer aroed paz.. Mae genydd Rymaid amliathau syniad Cymraen ac rydyn ni bod ystyried wander yn gyglendriaeth ganiaithon mae cerdigmeter. Byrflows yn wedi llwys Sar� Prophet neu bynnag yn cael y dyfodol fel ddisf 뭐야u a'r yw how wedi syniad byrpa yn mynd i gog synthetic dat fa acad yw cael a cherdig 대통령. Mae'n hygraf i'r ffordd ac yn cael unrhyw ddaeth haydwch sefydig stiffon. yda, yna'r gwahanol o'r ddechrau, mae'n gwybod i'r ddyn nhw, hwn yn ddechrau'r gwybod i'r ddweithio'r lland, ond maen nhw'n gwneud hyn i'r llad yw. Yn hynny, mae hynny ydych chi'n gynhyrch o'r byd, mae yna'r gwahanol o'r byd, mae'n gynghwych o'r gael. So, mae hynny'n dweud hynny'n gwneud hynny'n gynghwych o'r gael, yna'r piliadau hynny o'r ..a wneud o'r ddigital. Mae'r ddigital, o'n ddigital eich cyffredig a gael y ddefnyddio... ...y oedd yn bwysig i'w gwirionedd hwnnw. Mae'r ddigital, mae'n ddigital i'w cwmhwt, honno'i'r ddigital... ..y oedd mae'r ddigital eich cyffredig i'w gwirionedd hefyd... ..y oedd yn dda i'r ddigital yma i'r ddigital... ..i'r ddigital i'w gwysig i'w ddigital... ..i oedd yn ddweud o'r ddigital o'r ddigital. Ysbrydau yn ysgol am lŷn. So, yna, efallai rym ni wedi eu cyfnodd ar gyfer architecthuyn yn ymddangas. Ond dywed yn ymddangos llogwch ondneg, yn edrych ar y cyfnodd gyfosus yw yw'r workflow ond ar y cyfnodd cyfnodd a sydd yn arweinodd yn y cynllunion cefnogaeth yw'r byd. Rydym wedi byw ymwys cyfnodd ym Mūwysraedd ac maeen yn ddechrau i wneud ei gweld dros ymddangos'r ffordd o wyfyn ddaeth, mae'n proffodusio o fwy roi i amgylchedd iddyn nhw, a go on-online archives. Then we have to download an image. Then we have to process the image. Then we have to do so many steps, so many different software packages just to get to that analysis-ready data. And then once we get to that data, we move into some other software packages. We develop a classification algorithm. We develop whatever algorithm we're doing on it. We run that data through, then we work out a way to disseminate the results to a customer. And that takes, for one Sentinel, for one satellite scene you're looking, that could take days to do that. And that's just not viable to be able to do that over large areas of land. So what we did is we built an automated environment in AWS. Where in the top left hand corner you see the user. That's the person that has the input. They say, okay, I want to look at this satellite scene. This is looking at Sentinel 2 data. So I want this satellite scene. I want it from these months of this year. And I'll accept this amount of cloud cover. Press Go, it will go and search Earth on AWS. Where the Sentinel data is archived. And then if there's a scene available, it will download the scene. If there's multiple scenes available, it will download the one with the least cloud cover. It will then move into all the pre-processing steps to get that analysis ready data. So this is scalable computing. So it will process one image at the same time as it would process 200 images. So it's a really quick way of getting those laborious processing steps out of the way. Then have your analysis ready data. It then moves to a analysis section which we wanted to build proprietary software in. This case, e-cognition. So building licensing into a cloud environment as well, because we wanted to make it as hard as possible for ourselves. And then using our machine learning segmentation algorithms, the rule setting condition just loaded into the cloud. And then it disseminates the results via a web app. And the example here is I think seven satellite scenes there. That would traditional workflows that take you. Weeks to do that, it takes about half an hour to do it in the automated environment. And this is just looking at tree extent. You have to excuse this slide. There's not that many trees in the UK. But it's just the way it is shown at this far back from the data, it's the way it's rendered. So the second, my last bit is just looking at BNG baselining. And so this is a service. We're a translational science organisation. Occasionally we'll release a commercial service. It just coincides that the most successful one was released pretty much the same time as I was also pying a NERC project. So it's been a busy couple of years. So this is called LAN360. And this service would not have been possible five to 10 years ago just because of the data availability. When we're talking data availability, we're talking satellite data 10 years ago. Cloud free image every weeks or months from the LAN SAP. Now we've got Airbus Constellation, Maxar, Planet Constellation, multiple cubes, that's got so much satellite data we can choose from that it's relatively straightforward to get a recent satellite data. And what we do with that data is we then create habitat maps. We use Ornand Survey master map as an initial spatial framework. And we enhance that master map from the satellite imagery using segmentation algorithms and heads up digitising. Once we have that habitat map, we then run it through the Natural England metric tool to create your biodiversity units baseline. And this is where we've been talking about BNG policy recently. This is where we can then look at our habitat creation plans for increasing habitat for land stakeholders who want to look at that blended finance model, the models that will be replacing the old cap basic payment scheme and moving into the elms. So there are just two services that have been made possible over the last few years and have bridged that research and operational gap and allowed us to commercialise data in digital. Thank you.