00:00.40 archpodnet Ah, welcome back to episode one seventy nine of the archaeotech podcast and we are talking about a recent article in well I guess recent 2021 in archaeological advances from the essayas so go ahead and check the link to that in the show notes if you don't have the article. Ah, if you don't have access to the article. That's what we're here for so we're we're pretty much doing a deep dive on this one article for the entire episode. But again if you have access to it. It'd be great to go over and and take a look It's well written. So let's get into talking about what ips is we Paul just brought it up at the last minute in the at the end of the the. First segment. So now we're going to talk about it. But um, so first some I guess attributes of ips that that help it work when the other technology we meant the other technologies we mentioned don't work right? 1 of the things is obviously it doesn't rely on Gps satellites. This is not. A global positioning system. It's more of a local positioning system. They call it an indoor positioning system. But I think if if if projects like this archeological project. We talk about in segment 3 end up really using it and and other people in similar outdoor situations can use it. They may have to change the name because this was designed for indoor use which we'll get to but um, that will I guess I'll get to it right now according to my my handy bullet points here but ips is um was developed. Buy companies and one of the ones they talk about here is Marvel mind robotics. It's also linked in those show notes here. Um, it's developed by them to improve indoor robotics navigation and things like the tracking of customers and merchandise and stuff like that basically indoors. Ah, in big warehouses in you know, small warehouses. Whatever the case may be but tracking where things are and and just the current state of things and how they're fluctuating and moving and and also providing mapping systems for internal robots to move around. You might be thinking what the hell are we talking about but like Amazon for a while now has been using. Ah, robotic I guess ah pullers I don't know what you call them but robotic devices to pull and stock products off of shelves. There's still people in there. They're not fully replaced yet but to be honest, that's probably where it's heading but they've been using that and to do that accurately the the robotic devices need to have a really good. Map of what the area looks like and I know one of the common problems in robotics. These days is is devices mapping their surroundings in real time. It's getting better tesla vehicles do that I know that other other things are are getting better at doing that but mapping is 1 thing and interpretation is another so um. 02:37.67 archpodnet This technology was developed to help at least with the mapping portion of that and and help these things move around. 02:43.51 Paul Yeah, and what they do and this the system that they went with this Marvel mind system is just one of like you said ah any of a number of different competitors in this space now and they're mostly geared for like you said industrial applications and. Typically indoors applications though. I'm sure there are good ones. Outdoors. Um I'm sure they're being developed for various outdoor purposes. Um, you know again even with the inventory tracking. You could see things like on a car lot which is outdoor lumber yards things like that. Perhaps um. 03:00.31 archpodnet Right. 03:12.38 archpodnet Oh. 03:18.86 Paul But basically what they do is they set up a number of base stations and then your receiver whatever it is that you're tracking triangulates between those base stations and there are different ways of doing that and different technologies for for gauging the distance. But if you have 4 or more base stations. 03:29.73 archpodnet Right. 03:38.72 Paul Ah, you can get 3 dimensional position on something that can talk to all 4 of those stations. 03:45.78 archpodnet Yeah, exactly and the nice thing about this is it's entirely scalable. Um I was looking on the website earlier and I mean they can go up to something like twenty Thousand Square meters of continuous mapping which is insane if you just if you just link these all up together. And you put more of these devices out there. Um, and and you just it's scalable from I think the smallest area with the the 4 um ultrasonic sensors and then the 1 hedgehog which we haven't talked about any of that yet. But we'll get into it. Um was what was it like thirty Square meters or something like that and then if you start adding them on. It becomes scalable out from that point. 04:23.20 Paul Yeah, and so in concept it's a lot like GpsGnss right you have those satellites that your receiver can talk to or can receive the signals from and then do the math to figure out where it is in the world relation to to those satellites. 04:27.76 archpodnet Yeah. 04:34.41 archpodnet Um. 04:41.24 Paul So becomes the same sort of thing if you have those um you know the the base stations aren't moving unlike the ah the satellites which are but if you have the Receiver knows where those those base stations are you know where those base stations are it's just a math problem. 04:55.28 archpodnet Right. 04:58.35 Paul And then you can scale it out and like you said and and I've seen similar things being done for tracking using basically existing infrastructure for um for wi-fi networks within within buildings. You know so a lot of it is overlapping competing and I think that we're early on even though this article. 05:07.76 archpodnet Um, yeah. 05:18.34 Paul Talks about work that was done in 2019 and or a couple years past that. At this point I still get the feeling like this is a space with a lot of fairly rapid advancement and that things haven't glled yet that they're going to be different ones. Ah different methods. Ah. Different ways of relying on the signals I just mentioned wi-fi it would be bluetooth the ones that they discuss here uses ultrasonic sensors. Um, yeah, there are probably other things that we could use as well. I've got the feeling like we're going to start to see a congealing of. 05:39.40 archpodnet If. 05:53.97 Paul Certain things that work better in certain situations and certain ones that work better in other situations and will have a ah better sense and more robust set of tools that we as archaeologists can rip out of those warehouses and those stockyards and so on and and. 06:05.53 archpodnet No. 06:10.42 Paul Use them on our projects in the field. 06:11.96 archpodnet Yeah, and I mean you're totally right because I feel like this ultrasonic sensor one would not be super great in a lot of archeological environments and and you know purely because it relies on sound right? and there's almost I mean you might think that. You know archeology projects. We're archaeologists. We're out the middle and nowhere doing stuff but sometimes we're in the middle of a city sometimes we're out on the open plains. But it's like forty mile an hour winds while we're doing it and you know there's always people talking. There's usually music or something like that I mean we're we're never like super quiet so there would have to be some. Ah. 06:38.11 Paul M. 06:49.72 archpodnet I Guess some concessions taken in with these ultrasonic sensors because just to describe what this is really quick. It's really kind of a sonar from the sounds of it. My rudimentary understanding of it is kind of they're basically shooting out a signal and then that signal is is. 06:57.37 Paul Moving. 07:06.23 archpodnet Bouncing off of things and the time it takes that to come Back. Gives it a sense of distance and then they're using all of those sensors to basically triangulate points because they say well we hit this point from this one and this one and this one at these times and and then they're they're able to to map out the points. But again, that's relying On. That's relying on sound to come out and back and if you've got an interruption in that sound or other loud noises that are that are interrupting that sound process then then you've got problems. You know so. 07:35.88 Paul You know what? I'm curious about now that you mentioned that and this is not mentioned in the article but you know I was just saying that we're going to start seeing this or we already are seeing this in various industries I Wonder what's being done on construction sites. They're notoriously loud and they're usually. 07:48.87 archpodnet Um, right? Yeah, well even like. 07:52.59 Paul Semi if not fully outdoors so actually mining all leak out mining. Yeah, like when we worked last year I'm sure that that's being done to some extent the question and then becomes one of accuracy right? How close. 08:01.57 archpodnet Ah, yeah. 08:09.20 archpodnet Um, yeah. 08:11.88 Paul Do the how accurate do these have to be and what's the precision. You know this particular set um is in the range of two centimeters of accuracy which is not great, but it's definitely within the it's the same practically as most gnss systems good ones. 08:19.62 archpodnet Um. 08:30.40 Paul And it's um and it's not a whole lot worse than than total station serving which oftentimes is more accurate than is necessary for an archeological context. 08:32.97 archpodnet Ah. 08:41.45 archpodnet Well and I mean I Even wonder about the indoor situations where they're saying these were developed for how many how quiet is a warehouse you know what I mean like like how what what kinds of sounds and what level of sounds actually impact this thing negatively is what I would want to know because there's probably there's probably. 08:48.80 Paul E e. 09:00.66 archpodnet Some some pretty good reliability in the sound as far as different frequencies Probably don't expect it different Decibel levels probably don't affect it. So so maybe it's not as fragile as we think it is but um, yeah I mean that would be some serious concerns before you love this thing all the way across the world to an archeological site to give it a shot. 09:07.16 Paul Ah. 09:18.10 Paul But this particular technology as of 2019 and this is mentioned in the article it is susceptible to wind as a problem and that's yeah, like you said, that's not very good for us. It certainly would have been a problem in Nevada where those winds are just relentless. 09:19.46 archpodnet You know to try it out. 09:25.78 archpodnet Um, yeah, yeah, right right. 09:37.77 Paul And and even in Iraq and Lagosh where I was at the ah we we had windstors and strong winds quite regularly toward the end I couldn't fly the drone a number of days in road just because the winds they weren't strong strong but they were strong enough to be a problem. Um. 09:54.15 archpodnet Yeah, yeah, yeah. 09:57.70 Paul Yeah, so I don't know at what level that wind noise causes the problem with these ah systems and so I'm wondering you know again, what different kinds of sensor systems are being used and explored in order to overcome that. 10:09.13 archpodnet Well, the nice thing is unlike a $3000 Gps you may not know how to use or ah, you know 1 to $5000 drone. You don't know how to fly or operate or have a license for this starts at a pretty cheap price and you can probably just try it out and and and. And see what's going on and I think the price is listed in the article so it may be different now. But ah, three hundred and ninety Nine Euros for a starter kit which is about 434 us dollars. Um, right now as we're talking about this in may of twenty twenty two. And that's not a bad price for just getting the starter kit and just seeing if this thing will even do anything remotely like what you want to and I'm kind of like even if it's not super great, but it kind of gets part of the job done for that price I'll take it. You know what? I mean um, you know it's got to be reliable data that it's giving you but even if it's. 10:53.20 Paul E. 10:59.81 archpodnet Not able to reliably map like everything and and you're getting good good points over here and you can see that or you know I don't know I don't know what the thing would be but um, even partial success at that price would be would be a pretty good success compared to the cost that things usually are you know way up there. Um other cool things about this are. 11:16.12 Paul Yeah, and if its if it's ah the other cool thing is that if it's ah if it's scalable like they say it is that initial starter price means that you can expand it without breaking the bank right? and you can expand it to the needs of your project. 11:19.53 archpodnet Oh yeah, go ahead, Go ahead. 11:25.67 archpodnet Um, yeah. 11:33.86 Paul Probably still under a thousand dollars you know? ah for this part of the hardware and I will put a little asterisk on the on the cost. They also have 2 computers attached to it so you'd have to be able to take yeah so that's going to add. 11:34.47 archpodnet Yeah, yeah. 11:43.35 archpodnet Right? right. 11:49.94 Paul Roughly $2000 to the overall cost of the system. 11:51.66 archpodnet Yeah, so that's that's a good segue to just talk about the setup on this real quick before we run out of time on this segment because I want to spend the next one talking about the ah case study. So the some of the just quick and dirty and this is right from the article here. 12:02.64 Paul Mt. 12:07.29 archpodnet This uses a minimum again. This is a starter kit a minimum of four static beacons those ultrasonic beacons placed around the area that you want to start mapping and it uses ah a modem a radio frequency modem connected to a computer is the central hub. So. That's the first computer Paul's talking about so you got. A modem connected to that. That's basically picking up all these signals and then ah to me it started to get just a little bit confusing. Um one beacon is ah is a rover and it's called the hedgehog and I don't know why they called it a hedgehog maybe because it's got little intent is on it and stuff. But anyway, it's called a hedgehog and that thing. Um, receives info from the static beacons and sends it back to the computer so you've got the 4 static beacons and you've got the one moving beacon and I mean I saw some videos on their website. It's like they they had a version of this hedgehog that was like this little you know robotic thing on like 4 wheels that was running around I'm not sure how that would work on most. 13:01.21 Paul Yeah, yeah, yeah I was just going to say it I don't think that that's that that would work on any archaeological site but the video they have on their homepage is awesome. It doesn't remind me nearly so much as a hedgehog as the ah the mouse droids from Star wars. 13:03.45 archpodnet Archaeological sites. Ah no, no. 13:14.60 archpodnet Oh my god yeah, you're totally right? It's a hundred percent what it is ah oh that's awesome. How how appropriate since yesterday was star wars day may the fourth I know I know I know so. 13:25.46 Paul Oh boy. 13:31.41 archpodnet Anyway, so the hedgehog receives that info sends it back to the pc as well. The pc performs all the calculations on the data that is received and then also sends information back to the hedgehog to improve presumably. Its you know sense of space and where it's at and what it's already done. Um, and then all of that info was outputed to via Usb sorry via a Usb cable to a tablet and that's your typical output data standard and me data standard that you can see to you know Gps devices and stuff like that. Um, but it's a typical data stream that you can do stuff with not some proprietary. 14:07.37 Paul Right. 14:08.45 archpodnet Ah, some proprietary garbage that you can't do anything with um and then the the cool thing about all this I think I mentioned before the effective range for each beacon is thirty meters so if you want to go beyond thirty meters you need obviously more beacons but that's the whole thing. It's scalable. You can put as many beacons as you want on there. Probably up to a limit I think they they do have a limit of like 300 or something like that. But that's that's a lot. Um and then the whole thing can be georeferenced. So if you've got your you know your 1 ah submeter gps out there. You've got a control point somehow that you're using you can you can georeference this entire thing once you get the whole mapping. 14:29.34 Paul Yeah. 14:44.13 archpodnet Ah, the whole map being done so it sounds like they've done all that pretty well. 14:46.26 Paul Yeah, and I did want to to comment a little bit on those 2 computers that they use so the one that does all the processing I don't know what the minimum specs for it are but they said a windows or linux computer and so they used a fairly low-end windows computer. But I wonder if you could do it on something really low-end like a raspberry pi because that'll drop the price for that processor computer down to next to nothing I mean down to under a hundred buck so yeah and then the other one they're using it for um, for display. 15:06.42 archpodnet Oh yeah. 15:13.53 archpodnet Um, yeah. 15:21.32 Paul So it's taking that Andme output from the hedgehog and mapping it real time using Qgis which is awesome. But if you don't necessarily need that that real time display I Wonder if you could just hook up any kind of data logger to it and just. 15:23.52 archpodnet Okay, yeah. 15:36.90 archpodnet Oh yeah, probably. 15:40.21 Paul Capture it and then you would save yourself a lot of money and a lot of ah hassle and you wouldn't have to have a potentially fragile machine out there in the trenches around the ah the pickaxes and things. 15:45.54 archpodnet I mean I would I would imagine that would be the case right? I can't imagine you would need a real time display. That's just kind of a side benefit of it and when they're doing this for indoor purposes that totally makes sense and and seems legit but obviously for outdoor purposes. I Mean you're you're recording this almost for posterity and analysis later on not necessarily for current purposes. You know while you're in the field. So I mean it might be nice if you got the ability to see that display so you can make some corrections if you need to you know Remap an area or something I'm not sure how that works but you know that that might be a. 16:12.45 Paul Me. 16:24.25 archpodnet Ah, benefit. But yeah, either way it sounds like there's options. Yeah, and the the beacons are ruggedized. It sounds like so they're they're good for outdoor conditions and they have batteries in them but they can be powered from external sources as well. So it sounds like they've put some some thought into the hardware which is nice. 16:40.81 Paul Who. 16:42.49 archpodnet Ah, for some expandability it might have been designed for indoor use and they call it an indoor positioning system but it sounds like they've planned on some future applications which is really cool. 16:51.60 Paul Yeah, well I mean those ah those base stations you know, depending on the environment. Even if they're indoors. They may be subject to getting banged around you know I guess if your warehouse if you can securely mount them someplace. Well out of the range of everything you're you're good, but but if not. 17:00.93 archpodnet M. 17:11.54 Paul You know they might get bumped into by boxes or machinery or whatever So you'd probably want them somewhat ruggedized Anyhow, ah you know and warehouses and I don't know if this is one of their considerations. But yeah, they are big empty spaces and birds fly into them and. 17:13.45 archpodnet Um, yeah for sure. 17:28.75 Paul Nest up in the rafters and whatnot so you know, ah maybe if it's a bit cold and these are running a little warm because they're powered ah you know birds might have a tendency to make their nests on them and and poop all over them and and I think you'd want it a little bit ruggedized if that's going to be the case too. 17:29.51 archpodnet Yeah, yeah. 17:35.55 archpodnet Yeah. 17:44.30 archpodnet Seems legit. 17:45.30 Paul So even though they're interior for interior use I could think of some good reasons Why you'd want them ruggedized. Regardless. 17:49.76 archpodnet Yeah, indeed indeed so all right? Well one of my last little bullet points here mentions I forgot about this that when the wind is loud enough to be audible to humans it affects further distances than say 10 to twenty meters says wind errors are actually neg negligible at 10 to twenty meters from the beacon. So if you're if it's if it's you know, mapping something that close it actually isn't that susceptible to wind I'd kind of forgotten about that till I just good thing we take notes on these. Um I'd kind of forgotten about that. But when the wind is lot enough to be audible to humans so you can like hear that wind noise would be my guess. Um then ah. It it affects it at further distances than that 10 to twenty meters so all right? Well, let's go ahead and take another break and on the other side we'll talk about a cool case study that they did from a chocolithic site in Israel back in a minute.