WEBVTT 00:00.000 --> 00:13.400 Ready to go. Hello, everyone. So I have come to talk to you today about Kavesveg with a GIS 00:13.400 --> 00:17.400 sort of vibe. I'm assuming that I hardly anybody here knows anything about this subject. 00:17.400 --> 00:24.000 So this is very much interested. Apart from a few potholes with the audience, of course. 00:24.000 --> 00:34.000 So yes, I'm Wookie. I was a cave her before I was a free software nerd. So I started in 1987 when I went to university. 00:34.000 --> 00:44.000 And for about the last 25 years I've been running the UK Kavesveg group. So at least in theory I know something about the subject. 00:45.000 --> 00:54.000 So why do people survey Kavesveg? These cheerful looking people on the right hand side have just a skill of the whole load of passage and are looking very pleased of themselves. 00:54.000 --> 01:01.000 And that's all very nice, but if they don't survey it we don't know where it is because of course it's all hidden underground. 01:01.000 --> 01:13.000 So when we send in grant applications we call it science, but obviously mostly we all like maps here and you get pretty maps out of a lot of tedious caves of being. 01:13.000 --> 01:25.000 This is kind of 1990's vintage and this is a rather new Ashania colour. So they're just people who are used to this. The top is a side view elevation. 01:25.000 --> 01:34.000 One of them is an elevation and one is a plan. This is a mendic where all the rocks are about 45 degrees so there isn't really much difference between elevations and plans. 01:34.000 --> 01:49.000 So what is Kavesveg? How do you survey Kavesveg? We get on your passage. You basically take a line, pick a point, pick another point at the prominent point on the wall or a big rock and just get down the passage like that. 01:49.000 --> 01:58.000 It's extremely low tech and then you measure the distance to the walls and floor and roof and that's it. 01:59.000 --> 02:23.000 So in comparison to any end up with a center line which you can plot. So alright, how do we actually do that? So basically when I started at least it was done with an actual tape measure, a compass to measure the magnetic angle and a kilometer to measure the vertical angle and you scribble it down a bit of paper. 02:23.000 --> 02:31.000 And do a sketch round your of what you weather walls when and where the rocks are and that's the other thing. 02:31.000 --> 02:39.000 And then you type it up into a program so it makes a little bit of sense so you end up with a vertical point around here. 02:39.000 --> 02:54.000 So station 13 to 14 is 10 meters at 330 degrees minus 20 angle and then here's some notes on what the stations were and how far away the walls are. 02:54.000 --> 03:03.000 And then here's your crappy sketch into something that makes some kind of sense and then later on you produce a nice pretty survey to publish in a journal. 03:03.000 --> 03:13.000 Now you might imagine if you've seen people surveying by roads and the used teardolites and stuff on sticks and so why don't we do that underground that's been around for at least 30 40 50 years. 03:13.000 --> 03:29.000 Well the main reason is the conditions underground so first you haven't been caving I've just got a couple of minutes to kind of give you an idea what the conditions we have to go surveying in are. 03:29.000 --> 03:39.000 Results. 03:59.000 --> 04:23.000 I think I can. 04:24.000 --> 04:33.000 There we are that probably better. 04:33.000 --> 04:39.000 I wouldn't like to have to be embarrassed. 04:39.000 --> 04:44.000 I think I'm coming back. 04:44.000 --> 04:52.000 I've done it enough. 04:52.000 --> 04:59.000 It's a pity it's one of footwiser. 04:59.000 --> 05:09.000 I hope you can hear what they're saying it's quite stunning. 05:09.000 --> 05:14.000 It was just said I hope it's not full of water this is not full of water it could be worse. 05:14.000 --> 05:24.000 That's the same chap. 05:24.000 --> 05:39.000 I should thank this man for all this footage. 05:39.000 --> 06:04.000 It's not asking about that but it's not asking about that but it's going to get out of this. 06:04.000 --> 06:14.000 If you click the wrong thing it's quite a double click I think. 06:15.000 --> 06:31.000 Sometimes it's nice but often it isn't sometimes it's huge sometimes it isn't so we stick to extremely low tech surveying techniques there is of course no GPS underground so you can't do that either. 06:31.000 --> 06:46.000 It's not very accurate either so if you survey around at the same point there's often loops in case or you might survey to an entrance and then back to where you started if it's a through trip. 06:46.000 --> 07:00.000 So this is a bit of a problem we have to kind of do some distribute the air is basically around the loop so pretty early on people started writing software to do this because you don't want to do my hand. 07:00.000 --> 07:17.000 When I started it was programmable calculators out in the field so the earlier software I can find is by one Andy wanting to in 1982 it might be free software back in the days when people didn't put licenses on software they just kind of assumed people would use it. 07:17.000 --> 07:23.000 Unfortunately means technically it's proprietary but he's dead now so we can probably just. 07:23.000 --> 07:36.000 Nobody cares I mean it's probably a mainframe and then it was software for PCs and some max stuff and it's now I got started with service in 1991 I think it's safe to say that none of my code is still in the project. 07:36.000 --> 08:01.000 But that has been surprisingly popular 35 years later quite a lot of people still use service so we must have done something right about decade after that so that's just center line processing all of those tunnel and theory and so about decade later in 2000s we started using software to actually do the drawing part as well as the center line processing part and again both of those projects pretty much run it on now tunnels kind of dead. 08:01.000 --> 08:10.000 And there's loads more software I mean I'm only putting some of the projects on here so in the states as windows people. 08:10.000 --> 08:21.000 It's probably software we don't care about that but compass is still going to be going for years almost all of these projects are one or two dedicated people doing a project. 08:21.000 --> 08:32.000 And the quite region as well so walls was very popular with Americans going to Mexico visual topo also still running is pretty much the franca file caves are being software. 08:32.000 --> 08:42.000 If you run windows there's Jim and stuff as another French project which was in turbo Pascal and somebody has to. 08:42.000 --> 08:52.000 And there's new projects still coming on caveways quite recent QT based quite nice as he and there's even people still write non-free software don't want to wrong with them and hilariously. 08:52.000 --> 09:01.000 You can now rent your survey software by the month the Ariane projects aimed at divers who will have too much money so you know who cares. 09:01.000 --> 09:12.000 So this is all great except that we have like 30 different cave service software which all just invented a format and so none of them very few haven't taught to each other which means it's very difficult to move things between. 09:12.000 --> 09:20.000 Once you pick some software it's really quite difficult to move to something else often some of them will read other projects like service will read compass and walls data pretty well these days. 09:20.000 --> 09:25.000 Which is good because walls is dead so people are migrating and smaps is dead. 09:25.000 --> 09:34.000 There was a massive project where we set a certain grand project to translate everything to everything sort of GIS people you know you get everything feel thing but never really going anywhere. 09:34.000 --> 09:38.000 Cave converter is actually quite useful that does some formats. 09:38.000 --> 09:44.000 Service and theory and so close but not quite the same that a fairly stupid script will do the job. 09:44.000 --> 09:55.000 But eventually we realized like took a long time that there's real standards out there for 3D data and maybe we should use those instead and stop inventing our own crazy shit. 09:55.000 --> 10:00.000 So back to the actual instruments we started off with action compass and plano. 10:01.000 --> 10:10.000 But in about 2006 a Swiss guy called Beatheeb realized that you could take a standard like a distance measuring device. 10:10.000 --> 10:24.000 So all these gobbins insides out and replace them with some sensors for the magnetic and angled stuff but use the box and the screen and the buttons and so on and the laser module which is a really hard part. 10:24.000 --> 10:32.000 So and at the same time you also wrote some software to put on these windows, PDAs you could get in 2003 to draw. 10:32.000 --> 10:37.000 So Bluetooth the data to the other device and then draw the cave on the screen. 10:37.000 --> 10:44.000 And that was kind of kind of revolutionary. It took a while to catch on but it's more or less how most people do survey in these days because it's quicker and easier. 10:44.000 --> 10:52.000 And remove quite a lot of opportunities for errors all the shouting out of numbers when you read the tape measure and shout it down to somebody over the waterfall and they wrote down a bit of paper. 10:52.000 --> 10:57.000 And then you wrote times into a computer. Every one of those is an opportunity to make a mistake. 10:57.000 --> 11:08.000 So this is the devices you bought your like a device through a way that gobbins and put oops don't wrong edges again backwards. 11:08.000 --> 11:16.000 Put your little sticking board. Try this thing there this goes instead of that one. 11:16.000 --> 11:21.000 So that has all the magnetic and accelerometer sensors on and you put a non-magnific battery in. 11:21.000 --> 11:25.000 And this went pretty well for quite a long time cost you a few hundred pounds. 11:25.000 --> 11:29.000 It was just about cheap enough that people or at least groups could afford to do it. 11:29.000 --> 11:35.000 The problem of this game is that every time like a decide to bring out a new device you've got to redesign your gadget. 11:35.000 --> 11:44.000 And after this device they made it more compact and more integrated to make it cheaper which meant there was no way to talk to the display. 11:44.000 --> 11:47.000 It wasn't brought out to a separate bus you could just plug in. 11:47.000 --> 11:51.000 So basically it became impossible to maintain and you couldn't get the parts anymore. 11:51.000 --> 11:55.000 So for a while you couldn't get this dose. 11:55.000 --> 11:57.000 So it just do X's. 11:57.000 --> 12:05.000 Which was a bit of a problem because they're very valuable devices and people do of course drop them down pictures and fill them with water and generally stop them working. 12:05.000 --> 12:11.000 So the man in this room in fact quite a long time ago started going well actually we could make her own. 12:11.000 --> 12:19.000 So this 2008 was the Sheldon Attack Pony 2 which is the Compass and Plano part. 12:19.000 --> 12:23.000 That's the rest of easy bit and then you just use the separate electronic. 12:23.000 --> 12:27.000 Oh wow okay I sure I stand corrected. 12:27.000 --> 12:29.000 And yeah we use those they were good. 12:29.000 --> 12:35.000 And then there was the version 5 you could actually buy a laser module so you could make the whole device. 12:35.000 --> 12:40.000 So this is all open hardware unlike everything you've seen so well which is proprietary software and proprietary hardware. 12:40.000 --> 12:50.000 That was all right but the laser module wasn't great so it was a bit slow and the battery isn't as that long and it wasn't really good as any water anywhere. 12:50.000 --> 12:53.000 So now we're on to this thing which is actually pretty good. 12:53.000 --> 13:00.000 Better laser module and it does its own internal calibration which you have to do external and this is all marvelous. 13:00.000 --> 13:08.000 And because this is an open project people have started taking it and experimenting so this is basically the same hardware and software. 13:08.000 --> 13:16.000 In a better case with a lot more engineering as a result of which it cost nearly twice as much money but you know quite a lot more. 13:16.000 --> 13:20.000 And we're going to experiment with that next month to see if it's any good. 13:20.000 --> 13:27.000 Meanwhile the dystophing eventually so I spent an hour and a bus once with the at trying to persuade him that he should open up this software because this is tiny niche market. 13:27.000 --> 13:35.000 And he sold a thousand and 20 years whatever it is or 2000 and you know maybe if we could help him with a software if we were allowed to. 13:35.000 --> 13:40.000 But he wasn't having any of it but he did eventually publish the bomb after you couldn't build them anymore. 13:40.000 --> 13:48.000 So a couple of people updated them and did do some later builds and some Chinese guys now making basically dystop's in China. 13:48.000 --> 13:51.000 It's not particularly cheap but it's a nice bit of hardware. 13:51.000 --> 13:55.000 I saw one last week they do exist but at the moment you have to go to China to buy one. 13:55.000 --> 13:57.000 A little bit inconvenient. 13:58.000 --> 14:05.000 So yeah that's coming along nicely so that's all very very nice and that's what most people do but of course what we'd really want to do is like walk down the caves. 14:05.000 --> 14:12.000 Getting magic surveying from lighter obviously that's how it should work so we've been sort of desiring this for decades. 14:12.000 --> 14:29.000 But I was around about 2014 I think this thing came out which is not Australian device which is so that's a 270 degree laser planar gadget mounted on top of an emotional IMU on a spring which is kind of weird you go why is it on a spring. 14:29.000 --> 14:35.000 Well actually that's quite important part of the thing because IMU is a shit and forget where they are at every 30 seconds. 14:35.000 --> 14:50.000 So in order to have that track basically as you flop the thing backwards and forwards on its spring the laser keeps coming back to the same point so you can keep recalibrating your IMU every couple of seconds which keeps it more or less in train. 14:50.000 --> 15:01.000 And yeah it's quite good we took it caving it's surveys things nicely the problem is that the nine was it 14,000 pound a buy it was you have to pay for the data to be remotely processed by the cubic meter. 15:01.000 --> 15:13.000 So it was extortion so only only rich archaeologists and professionals used it we tried to make a run by buying the two devices for about five grand turns out the sums are genuinely difficult. 15:14.000 --> 15:22.000 So yes more like that so this is actually a proper expensive LIDAR ridial things like 25,000 dollars or something. 15:22.000 --> 15:31.000 But with that you can go and survey big case so this is there's actually a project to survey all the massive caves in the world with the fancy LIDAR. 15:31.000 --> 15:37.000 So they'd be dragging it around so you get all these laser points which show you which basically make a shape. 15:37.000 --> 15:45.000 So that's radically different data to what we normally use now the bit of look I can show you. 15:45.000 --> 15:51.000 So then you can take your point cloud and then color the points which is not what we want. 15:51.000 --> 15:56.000 With actually I remember I'm a bit of a thing tonight. 15:57.000 --> 15:59.000 Sorry about this. 15:59.000 --> 16:02.000 I was hoping if I clicked on it you would take it there. 16:02.000 --> 16:05.000 Then mine it's right there with me. 16:05.000 --> 16:09.000 Should be a window I can't be prepared earlier. 16:09.000 --> 16:29.000 So if you take photos you can color all your laser points with the right point on the photos. 16:29.000 --> 16:35.000 So now you get this quite funky so this is a showcase of Germany and you can spin around and look at things. 16:35.000 --> 16:41.000 So you get even though it's still just a point cloud you actually get kind of some idea of what the things like. 16:41.000 --> 16:43.000 And this is a technology called Poetries. 16:43.000 --> 16:48.000 So the point you've got millions of millions of millions of points and it's really quite difficult to manage. 16:48.000 --> 16:54.000 You want a lot of points right next to you but you don't want all the billions of points off in the distance because it would just be too slow. 16:54.000 --> 17:02.000 So Poetries is a kind of binary chop spatial thing which lets you view data like that successfully. 17:02.000 --> 17:08.000 So you can then stick your photos around the edge just in the 3D model and make really quite snazzy pictures. 17:08.000 --> 17:12.000 That's quite difficult to make stuff like this as a whole load of work involved but it's very nice. 17:12.000 --> 17:15.000 So back to what my models are doing. 17:15.000 --> 17:21.000 The little PDAs you could get in 2003 eventually became an obtainium and the old batteries died. 17:21.000 --> 17:24.000 But obviously everybody had lots of Android devices. 17:24.000 --> 17:31.000 So this sexy topo on the left is basically a rehabilitation of pocket topo for Android. 17:31.000 --> 17:33.000 And then this is topo droid. 17:33.000 --> 17:36.000 So those are the two free software programs that nearly everybody uses. 17:36.000 --> 17:39.000 This topo droid is basically theory on Android. 17:39.000 --> 17:43.000 The difference is that here you're just drawing lines around a center line. 17:43.000 --> 17:49.000 It's just a picture whereas in this you're putting metadata in saying this line is a wall on this line is a pitch edge and this line is a whatever. 17:49.000 --> 17:52.000 So you're doing more detail underground which takes a bit longer. 17:52.000 --> 17:56.000 So it kind of depends how miserable your caving is which is where you want to go. 17:57.000 --> 18:01.000 There's forever isn't no different any else software yet for cave surveying. 18:01.000 --> 18:04.000 There's a few proprietary things but again we don't care about that. 18:04.000 --> 18:08.000 So cave surveys have been looking like this for a long time. 18:08.000 --> 18:15.000 So in about 100 years these aren't fundamentally very different. 18:15.000 --> 18:18.000 And some people think this is bad. 18:18.000 --> 18:23.000 So Julian thought I know what we wanted to do is draw in 3D. 18:23.000 --> 18:31.000 Stop making it's basically all these surveys are a 2D flattening onto a plane to make a map which is fine. 18:31.000 --> 18:38.000 But you know caves are very 3D things and it's quite difficult to get them onto a plop piece of paper and still be useful. 18:38.000 --> 18:43.000 So using the god-o game engine you can take your point cloud, stick it up and then kind of draw rounds. 18:43.000 --> 18:45.000 Just the same way we do a street map. 18:45.000 --> 18:51.000 You draw down the road and then you draw around the edge of a field and say that's a field similar kind of technology. 18:51.000 --> 18:58.000 And you end up with so you get away from the billions of dots which your brain's quite good at interpreting when you spin it round. 18:58.000 --> 19:01.000 But you can't really print it out in a useful way. 19:01.000 --> 19:08.000 So as this cuts it down to a plausible number of triangles so you can actually get kind of useful maps of it. 19:08.000 --> 19:12.000 You need a VI headset to draw it but it's kind of fun. 19:12.000 --> 19:15.000 So that's quite a cool idea. 19:15.000 --> 19:26.000 So just go back a bit when we started back in 1989 we had this 10-ology which is basically being able to look at your caves. 19:26.000 --> 19:30.000 So here's the Royal Centre lines that you drew and you get a nice spinny cave. 19:30.000 --> 19:37.000 So that's 140 kilometers of cave and I can move it around quite quickly and I can look at well the entrances are. 19:37.000 --> 19:43.000 All these ones down the bottom idiots who can't write down the altitude of their phones. 19:44.000 --> 19:47.000 So zero obviously. 19:47.000 --> 19:52.000 And we can put some terrain on so you can possibly see I get this right. 19:52.000 --> 19:57.000 This cave can turn the entrance off because that's just confusing. 19:57.000 --> 20:04.000 This cave is kind of in a ridge here and there's a lake down the bottom. 20:04.000 --> 20:07.000 There was a little bit there you can see the lake here. 20:07.000 --> 20:12.000 So this is this is cave is writing their own GIS it's kind of primitive GIS. 20:12.000 --> 20:19.000 All you get is terrain data and your centre lines but it works and we've been using that for 20 years. 20:19.000 --> 20:28.000 This is slightly snazzy a version of the same thing so now we can drop a map onto the terrain data model and have the caves underneath. 20:28.000 --> 20:35.000 So this is a mendip in the UK so that's some you can see the caves and you can kind of see here. 20:35.000 --> 20:41.000 There's a valley with a cave underneath and in fact we're going to drive this. 20:41.000 --> 20:49.000 We can look underneath and go yeah there's caves under the under the ground. 20:49.000 --> 20:53.000 Which is of course my point about the whole z axis thing maps are not flat. 20:53.000 --> 20:55.000 Flat map for software that can't cope. 20:55.000 --> 20:59.000 It seems everything's on the same level is a bit of a problem for caveers. 20:59.000 --> 21:05.000 So I don't do that to be fair mostly people this problem appears to be largely fixed. 21:05.000 --> 21:11.000 So we discovered after about 25 years that there was GIS software we could use to be fair. 21:11.000 --> 21:15.000 It's only been available to mere mortals relatively recently. 21:15.000 --> 21:22.000 So rather than us adding more and more GIS features to our crappy caves software which is very good at processing centre lines. 21:22.000 --> 21:28.000 Maybe we should process centre lines then import the cave data into more generally capable software. 21:28.000 --> 21:35.000 Where it becomes trivial to add contours and satellite images and terrain data models and all that stuff. 21:35.000 --> 21:37.000 So that's what we started doing. 21:37.000 --> 21:39.000 And obviously you can do this on your phone as well. 21:39.000 --> 21:47.000 So on MSM and you can have all your cave entrance locations and then GPS tracks so you don't get lost on the mountain. 21:47.000 --> 21:49.000 Well hopefully. 21:49.000 --> 21:56.000 And we just started doing this using Qfield so in field stuff for recording new caves. 21:56.000 --> 21:59.000 Because people are again especially on the expedition I want it's full of students. 21:59.000 --> 22:02.000 People turn up they have no fucking clue what they're doing. 22:02.000 --> 22:05.000 So they produce very poor data. 22:05.000 --> 22:12.000 So as if you give them a little feel to fill in we've got a fighting chance that they will actually take a photo of the entrance and take a GPS point with an altitude. 22:12.000 --> 22:14.000 And so on. 22:14.000 --> 22:19.000 So if anyone's actually tried that I would be interested to hear of your experiences. 22:19.000 --> 22:24.000 So we collect an off-loc cave data and we have to manage it. 22:24.000 --> 22:28.000 So the thing called cavemaps.org really useful projects. 22:28.000 --> 22:31.000 Probably been there for 12 more years now. 22:31.000 --> 22:33.000 Someone took all the surveys. 22:33.000 --> 22:39.000 So they're all flat surveys in the UK drawn in the kind of 50s, 60s, 70s, 80s. 22:39.000 --> 22:41.000 All paper surveys. 22:41.000 --> 22:44.000 So he went through, collected them all, scanned them all. 22:44.000 --> 22:48.000 So you can now find, so it's called cavemaps.org because of it covered the whole world. 22:48.000 --> 22:51.000 It actually covers Yorkshire. 22:51.000 --> 22:53.000 It could cover more. 22:53.000 --> 22:55.000 Anyway, it does a very good job of Yorkshire. 22:55.000 --> 22:59.000 So you find you get how many surveys there are and you can get a list. 22:59.000 --> 23:04.000 And then this is the 1949 survey of Mostell Caverns. 23:04.000 --> 23:10.000 And then we can have the 1989 40 years later, which if you look actually matches at reasonably well. 23:10.000 --> 23:15.000 They found some more cave, but it wasn't terrible to start with. 23:15.000 --> 23:18.000 So that's all very good. 23:18.000 --> 23:23.000 The problem is that those surveys, people did all this cave surveying. 23:23.000 --> 23:27.000 I've tried to explain to you and they drew it up onto a nice paper and they published it in a journal. 23:27.000 --> 23:30.000 And then they threw away the notes because it was finished. 23:30.000 --> 23:36.000 And that's very annoying now when we want nice 3D models of all our caves with the actual central lines. 23:36.000 --> 23:39.000 You go back to these people and go, where's the original data? 23:39.000 --> 23:40.000 They get all I don't know. 23:40.000 --> 23:41.000 Am I throwing it away? 23:41.000 --> 23:43.000 Maybe it's in the loft. 23:43.000 --> 23:46.000 Oh, they've died because it was 40 years ago. 23:46.000 --> 23:51.000 So we can survey them again, but for some of these caves, that's pretty hard work. 23:51.000 --> 23:59.000 So in 2005, the UK case of our group set up the cave registry to try and at least preserve the original data. 23:59.000 --> 24:02.000 So people wouldn't just loot, they had somewhere to send it. 24:02.000 --> 24:06.000 If somebody has died and they found all these useless cave notes, go, yes, please send it to us. 24:06.000 --> 24:07.000 We can deal with that. 24:07.000 --> 24:15.000 So we've been collecting data for 20 years for quite all the UK and various other projects around the world. 24:15.000 --> 24:19.000 Actually, people used to be very proprietary about their data. 24:19.000 --> 24:22.000 And they've gone through a lot of trouble to get it and they've published a map with it. 24:22.000 --> 24:24.000 And they didn't want to give it to anybody because they might. 24:24.000 --> 24:27.000 I don't know, published their own maps or look at it or something. 24:27.000 --> 24:33.000 So we've had definitely, honestly, nearly everybody has understood that this is a very niche thing. 24:33.000 --> 24:36.000 And there's no point keeping it to yourself. 24:36.000 --> 24:39.000 And we should share. 24:39.000 --> 24:44.000 There's still a few holdouts, they're annoying, but things are much improved. 24:44.000 --> 24:48.000 Now, I haven't been very long to cover this, but I just thought also to talk about, 24:48.000 --> 24:51.000 especially exploration projects, long-term ones. 24:51.000 --> 24:56.000 You collect a lot of data over a long time and you get new people who don't know what they're doing every year, 24:56.000 --> 24:57.000 or as we do. 24:57.000 --> 24:59.000 And also, obviously, formats change. 24:59.000 --> 25:01.000 You know, it used to be a compass and clienteau, 25:01.000 --> 25:04.000 and increasingly we come to LiDy data. 25:04.000 --> 25:11.000 And some of it, these days you tend to get a file of an Android device, 25:11.000 --> 25:16.000 as you know, original data rather than some scribbly notes to scan. 25:16.000 --> 25:18.000 So keeping track of all that. 25:18.000 --> 25:20.000 And the problem is that people go back to a cave 25 years later, 25:20.000 --> 25:23.000 and they want to find the end of the survey and carry on. 25:23.000 --> 25:26.000 And they need to be able to find, well, where was it? 25:26.000 --> 25:27.000 What was it? 25:27.000 --> 25:28.000 How do I find the note? 25:28.000 --> 25:29.000 What did people write down? 25:29.000 --> 25:33.000 So people come back to base with some crappy notes like this, 25:33.000 --> 25:38.000 and hopefully they write up a logbook entry, which might have a diagram, 25:38.000 --> 25:40.000 some hard to read handwriting. 25:40.000 --> 25:47.000 Obviously, increasingly getting anyone under 25 to write on paper is hopeless these days. 25:47.000 --> 25:50.000 So you have to give them a web interface to type into. 25:50.000 --> 25:52.000 You know, it's fine, we can do that. 25:52.000 --> 25:58.000 But you know, organizing all this stuff over decades is a really interesting problem. 25:58.000 --> 26:01.000 It's very heterogeneous. 26:01.000 --> 26:04.000 So anyway, so we have a big Django website, 26:04.000 --> 26:08.000 which tries to do things like like have all the people and all the trips they did, 26:08.000 --> 26:10.000 and which surveys they produce. 26:10.000 --> 26:11.000 Is there a survey for this trip? 26:11.000 --> 26:13.000 And is there a logbook right up for this trip? 26:13.000 --> 26:17.000 And then the colours match if you've got people on the same trip on the same day, 26:17.000 --> 26:19.000 so you can tell they're on the same trip. 26:19.000 --> 26:21.000 So that's sort of cross referencing is really useful. 26:21.000 --> 26:24.000 And then you can find the original survey data from where this one is. 26:24.000 --> 26:27.000 Here's a 2018 one. 26:27.000 --> 26:31.000 And we have an index to all the caves people have found, 26:31.000 --> 26:36.000 which should taste a little webpage, which should contain the info that's actually there. 26:36.000 --> 26:38.000 So ideally there will be a survey, 26:38.000 --> 26:40.000 and some sort of description telling you what this is. 26:40.000 --> 26:44.000 There's an awful lot of crappy little caves in between the massive interesting ones, 26:44.000 --> 26:49.000 but you don't want to keep finding the same entrance and exploring. 26:50.000 --> 26:53.000 Have it to get a rope, often go down. 26:53.000 --> 26:54.000 Checkers and go anywhere. 26:54.000 --> 26:55.000 Hopefully someone took a photo with the entrance, 26:55.000 --> 26:58.000 so when you get there you can tell that you're at the right damn cave, 26:58.000 --> 27:00.000 but they often didn't, 27:00.000 --> 27:05.000 especially again 30 years ago before GPS was much easier to lose your caves. 27:05.000 --> 27:07.000 Or at least we have records of, 27:07.000 --> 27:09.000 okay, we just don't know where it is. 27:09.000 --> 27:12.000 And then you find them again, they get a new number. 27:12.000 --> 27:15.000 And eventually 12 years later you work out is the same thing. 27:15.000 --> 27:17.000 It's all good fun. 27:17.000 --> 27:19.000 Anyway, that is basically it. 27:19.000 --> 27:21.000 I will stop talking. 27:21.000 --> 27:23.000 I could obviously go on for hours, 27:23.000 --> 27:25.000 but that's enough for now. 27:25.000 --> 27:27.000 You've got the general idea, I hope. 27:27.000 --> 27:29.000 If anyone wants to talk to me about 27:29.000 --> 27:34.000 the joys of using Q field and merging maps and merging big datasets, 27:34.000 --> 27:36.000 or managing datasets, 27:36.000 --> 27:38.000 I am interested, and there's other people around who are. 27:38.000 --> 27:39.000 Thank you. 27:40.000 --> 27:41.000 Thank you. 27:47.000 --> 27:49.000 He's got any questions. 27:53.000 --> 27:55.000 I really watch too many say-fame always, 27:55.000 --> 27:57.000 but what about drones, 27:57.000 --> 27:59.000 and so things that do it for you? 27:59.000 --> 28:00.000 It's a good point. 28:00.000 --> 28:02.000 I didn't mention that people have tried drones on the ground. 28:02.000 --> 28:04.000 They work surprisingly badly, 28:04.000 --> 28:06.000 so firstly there's no GPS, 28:06.000 --> 28:09.000 and a lot of them rely on that to do a hover. 28:09.000 --> 28:12.000 So we took a very expensive drone underground in Ireland. 28:12.000 --> 28:14.000 The other problem is they suck themselves into the walls, 28:14.000 --> 28:16.000 because of the Venturi effect. 28:16.000 --> 28:18.000 So as soon as you get a bit closer to the wall, 28:18.000 --> 28:20.000 it goes, oh, and smacked into the wall, 28:20.000 --> 28:21.000 and falls down a shaft. 28:21.000 --> 28:24.000 And now your expensive drone is down a big hole. 28:24.000 --> 28:26.000 So we did that. 28:26.000 --> 28:27.000 It bounced off the wall, 28:27.000 --> 28:31.000 and we managed to catch it before the £1000 drone fell down the hole, 28:31.000 --> 28:33.000 but we decided we should stop at that point and give up. 28:33.000 --> 28:35.000 So there's the sort that have like bands around, 28:35.000 --> 28:37.000 and that kind of works a bit better, 28:37.000 --> 28:39.000 and you can, but it's a very expensive, again, 28:39.000 --> 28:42.000 I think, Swiss drone designed for inspecting tunnels, 28:42.000 --> 28:44.000 which has a carbon fiber shell round, 28:44.000 --> 28:47.000 and is designed to work without a GPS signal. 28:47.000 --> 28:51.000 But I've seen them use that on a national geographic glassier thing, 28:51.000 --> 28:53.000 but again, it's 20 grand, 28:53.000 --> 28:55.000 it's not accessible to me in mortals. 28:55.000 --> 28:56.000 It could be done, I think. 28:56.000 --> 28:58.000 Now they've got a lot cheaper. 28:58.000 --> 29:01.000 There's probably software for just not doing the GPS thing, 29:01.000 --> 29:03.000 but it's difficult to fly drone underground. 29:03.000 --> 29:05.000 You can't see it to the other problem. 29:05.000 --> 29:08.000 As soon as it's more than 20 yards away, 29:08.000 --> 29:11.000 or if you're trying to go up there to look to see if I think, 29:11.000 --> 29:13.000 I mean, we're really, really useful for things 29:13.000 --> 29:15.000 when you can see there might be a hole in the ceiling, 29:15.000 --> 29:16.000 40 meters up, 29:16.000 --> 29:18.000 and it's an unbelievable amount of work 29:18.000 --> 29:20.000 to go and get yourself up there. 29:20.000 --> 29:22.000 People do that. 29:22.000 --> 29:23.000 So no, yes, it is interesting. 29:23.000 --> 29:27.000 People haven't had much success so far. 29:28.000 --> 29:31.000 So my question is, how do you assess the quality of data 29:31.000 --> 29:32.000 coming into the K registry? 29:32.000 --> 29:34.000 Because obviously you've got pen and paper, 29:34.000 --> 29:36.000 and you've got later data, 29:36.000 --> 29:37.000 and you've got later data. 29:37.000 --> 29:40.000 Like how do you know what data's good enough to press? 29:40.000 --> 29:44.000 So, what the answer to the question is, 29:44.000 --> 29:45.000 we don't care. 29:45.000 --> 29:47.000 So each project manages, 29:47.000 --> 29:48.000 we give them repository, 29:48.000 --> 29:50.000 and they can do what they like in there. 29:50.000 --> 29:52.000 The object is to not lose the data, 29:52.000 --> 29:53.000 right? 29:53.000 --> 29:55.000 So we don't care what format they keep it in, 29:55.000 --> 29:57.000 how good it is, 29:57.000 --> 29:59.000 what could care they take of it, 29:59.000 --> 30:00.000 or any of that stuff. 30:00.000 --> 30:03.000 We just give people basically a file store. 30:03.000 --> 30:05.000 Now, obviously, 30:05.000 --> 30:08.000 it is important if people are sharing data, 30:08.000 --> 30:09.000 or other people's, 30:09.000 --> 30:11.000 so we do have a grading system for cave surveys, 30:11.000 --> 30:14.000 which is, again, designed 30 years ago, 30:14.000 --> 30:17.000 in the age of obstacle instruments, 30:17.000 --> 30:19.000 and really doesn't need an update actually, 30:19.000 --> 30:21.000 because things are changing different now. 30:21.000 --> 30:25.000 The new digital instruments are definitely better. 30:25.000 --> 30:28.000 The accuracy of the instrument is similar, 30:28.000 --> 30:29.000 it's probably a bit better, 30:29.000 --> 30:30.000 but the main thing is, 30:30.000 --> 30:32.000 there are dramatically fewer mistakes, 30:32.000 --> 30:34.000 just because you're Bluetooth in the data, 30:34.000 --> 30:37.000 and nobody's writing it down three times 30:37.000 --> 30:39.000 on muddy bits of paper. 30:39.000 --> 30:43.000 Do you export the data to open street work? 30:43.000 --> 30:44.000 I'm not doing that, 30:44.000 --> 30:45.000 but people are, 30:45.000 --> 30:46.000 so I did mention the slides, 30:46.000 --> 30:47.000 and there's a lot of cave entrances 30:47.000 --> 30:50.000 in the UK and France are on open street map. 30:50.000 --> 30:51.000 It's really, really useful. 30:51.000 --> 30:53.000 It used to spend hours wandering around mountains, 30:53.000 --> 30:54.000 and looking for entrances, 30:54.000 --> 30:56.000 which you knew were over here somewhere, 30:56.000 --> 30:58.000 and now you just take your phone with you, 30:58.000 --> 30:59.000 without, 30:59.000 --> 31:00.000 okay, I don't know, 31:00.000 --> 31:01.000 okay, venture is marvelous. 31:01.000 --> 31:03.000 So, yeah, I think we could do more of that. 31:03.000 --> 31:05.000 It's not, I don't think it's been very organized so far, 31:05.000 --> 31:07.000 and it's just people putting entrances. 31:07.000 --> 31:09.000 I did remove one last week, 31:09.000 --> 31:11.000 because Breitweight Hole has a cave entrance, 31:11.000 --> 31:13.000 and there is an entrance in Breitweight Hole anymore. 31:13.000 --> 31:15.000 There used to be 30 years ago, but it fell in. 31:15.000 --> 31:18.000 So, yeah. 31:19.000 --> 31:21.000 We could definitely do more, and it's good. 31:21.000 --> 31:24.000 There's a slight problem that Americans hate the idea. 31:24.000 --> 31:27.000 There's this long-standing problem that Americans protect their caves, 31:27.000 --> 31:30.000 by not mentioning where the entrances are, and keeping them secret. 31:30.000 --> 31:32.000 Everyone else in the world has horrible caves, 31:32.000 --> 31:34.000 that don't really have a problem with people going in 31:34.000 --> 31:37.000 and shooting the bats with shotguns and stuff. 31:37.000 --> 31:40.000 So, so we don't have to worry about it. 31:40.000 --> 31:45.000 So, Europeans just published entrance locations by large. 31:45.000 --> 31:48.000 And there was a big fight on Wikipedia about 12 years ago 31:48.000 --> 31:52.000 when Americans tried to insist that entrance locations were protected data. 31:52.000 --> 31:53.000 Nobody could put them up. 31:53.000 --> 31:54.000 And you go, fuck off. 31:54.000 --> 31:55.000 It's all grown in. 31:55.000 --> 31:57.000 You can keep yours, can't we? 31:57.000 --> 31:59.000 They were trying to keep it out of the template. 31:59.000 --> 32:00.000 That was it. 32:00.000 --> 32:02.000 Very annoying. 32:02.000 --> 32:05.000 I was thinking of the actual passage data as well. 32:05.000 --> 32:08.000 Obviously, provided it doesn't get interrooting engines. 32:08.000 --> 32:13.000 Interesting question how useful that is to everybody. 32:13.000 --> 32:16.000 We could have got loads of passage data to put in. 32:16.000 --> 32:19.000 I have done a little one in France where you could just walk in one entrance. 32:19.000 --> 32:20.000 It was easy cave. 32:20.000 --> 32:21.000 You could come out another one. 32:21.000 --> 32:22.000 So, I put a path through. 32:22.000 --> 32:23.000 I see somebody has removed it again. 32:23.000 --> 32:25.000 I check that's night. 32:25.000 --> 32:27.000 So, yeah. 32:27.000 --> 32:29.000 Interesting question. 32:29.000 --> 32:30.000 What people would? 32:30.000 --> 32:33.000 I think we should ask those same people what they think. 32:33.000 --> 32:35.000 We've got loads of data. 32:35.000 --> 32:38.000 Rooting in caves is a thing which is not solved. 32:38.000 --> 32:40.000 Because often there's more than one way to get there. 32:40.000 --> 32:45.000 And it's highly variable how difficult it is and how much equipment you need. 32:45.000 --> 32:48.000 We could actually, for that big complicated system I showed you, 32:48.000 --> 32:51.000 we could really do with some rooting software that we kind of estimate. 32:51.000 --> 32:54.000 If I want to get from it, because people are coming back 40 years later. 32:54.000 --> 32:57.000 We knew how to get there when we found it. 32:57.000 --> 33:00.000 But now, some 20-year-old going, I want to get to here. 33:00.000 --> 33:01.000 How the hell do I do it? 33:01.000 --> 33:03.000 It's actually got a hard problem. 33:03.000 --> 33:13.000 Is there a consensus on formats and standards now? 33:13.000 --> 33:16.000 Or is that still just unsolved? 33:16.000 --> 33:17.000 No, there isn't. 33:17.000 --> 33:19.000 But there's fewer bits. 33:19.000 --> 33:23.000 There's a certain amount of, they used to be like 50 cave programs. 33:23.000 --> 33:26.000 And now there's about 10 that people are still using, I think. 33:26.000 --> 33:28.000 So, it's less of a problem. 33:28.000 --> 33:31.000 But, I mean, in terms of the data formats. 33:31.000 --> 33:37.000 I think, you know, because there's the 3D format that, like, 33:37.000 --> 33:40.000 Therian and Servex share, they both output that. 33:40.000 --> 33:46.000 And there's a plugin, which is in QGIS, which will import it. 33:46.000 --> 33:48.000 So that basically just works. 33:48.000 --> 33:52.000 I don't know if this QGIS plugin is for anybody else's formats. 33:52.000 --> 33:55.000 There might be. 33:55.000 --> 33:58.000 So, yeah. 33:59.000 --> 34:04.000 So, I think there is people understanding that we shouldn't be using our own crazy formats. 34:04.000 --> 34:06.000 Any more than we absolutely have to. 34:06.000 --> 34:12.000 So, I think just the sense of line processing part will probably remain specialist formats. 34:12.000 --> 34:14.000 And feel of them. 34:14.000 --> 34:17.000 But everything else, we're just used to standard things. 34:17.000 --> 34:20.000 You know, GJs and Shakefiles and all that. 34:20.000 --> 34:21.000 Malaki. 34:21.000 --> 34:22.000 Thank you. 34:22.000 --> 34:26.000 I was just wondering, is the data accurate enough that you can tell whether you could 34:26.000 --> 34:30.000 physically squeeze through certain openings and whether you can't throw others? 34:30.000 --> 34:32.000 Yeah, pretty much. 34:32.000 --> 34:38.000 Generally, you can tell from a certain, I mean, the surveys are annotated with things like squeeze and two tight. 34:38.000 --> 34:40.000 And obviously, it's subjective. 34:40.000 --> 34:44.000 How small a person and how enthusiastic are they can you find? 34:44.000 --> 34:47.000 You know, it's an excellent sport for tiny women. 34:47.000 --> 34:50.000 Otherwise often disadvantage in sport. 34:50.000 --> 34:55.000 So, yeah, generally the surveys are, you can kind of tell what's. 34:56.000 --> 34:58.000 Practical. 34:58.000 --> 34:59.000 Thank you.