00:00.00 archpodnet Welcome back to episode one seventy seven of the architect podcast and as I mentioned it's just me today. Paul is out strolling around in Iraq looking for fun and interesting archeology. So again, we'll talk to him when he gets back about that. But we're here with Daniel talking about robotics and archeology and. You mentioned in the in the last segment Daniel I mean you had a a hard time which I totally understand I like just defining what archeology means so you can say I want a robot to do archeology and man we talked about that on the archeology podcast network and the crm podcast all the time like what is. What does archeology actually mean you know I mean is it is it research. Is it digging. Is it walking around on the ground and looking at stuff is it writing reports I mean what is that's a hard definition but I was also wondering why you were saying that. How do you find? How do you? Define robot. Right? Like because a robot could conceivably just be an intelligent computer program that does things for you. What about drones are drones running on an automatic program that you set up for them or maybe even one that says hey I want this area surveyed you figure out how to do it. Um is that also a robot like in in your. But first I want to hear you define robots for the purposes of your thesis but then maybe also define robot for what you're trying to achieve because some people might be imagining. You know a silon from battlestar galacica or something like that you know arms and legs and and you know just using a shovel to dig a hole. So. 01:25.62 Daniel Carvalho This is so. 01:28.44 archpodnet So how did you Define Robot for this study and and how were you defining robot for exactly what you're looking for. 01:33.55 Daniel Carvalho Okay, and this is another complex question because roboticists probably have this problem too The meaning of robot is is not pleasant when it's it came from ruhr. 01:41.43 archpodnet Yeah, and probably. 01:53.29 Daniel Carvalho And a marvelous piece of literary work and it means labor force ah work forced work. So it's difficult to say what is a robot today because many robots. 02:07.86 archpodnet Um. 02:11.90 Daniel Carvalho Today lack a physical form and that was like a major definition in classical robotics that robotics is something that we can. Induce a kind of function but that something was material but now we are on an entirely new dimension. We have robots that do not exist in our. 02:32.76 archpodnet Right. 02:49.70 Daniel Carvalho Present life as objects we have artificial intelligence we have ah drones that are a kind of eyebrid because drones and I'm glad you asked that drones are the main robot that we view today. And they are a hybrid. They have a physical body but they have a brain a set of algorithms that we can program to do certain functions. They fly they take photos. 03:10.76 archpodnet M. 03:23.53 Daniel Carvalho But they can identify objects too. So They have computer vision and so I think the the most useful definition that we can have of a robot concerning humanities and archaeology is. And machine and that can function ah with a set of instructions. Ah the that set of instructions being made by a human. 03:50.45 archpodnet Okay. 03:57.50 archpodnet Um. 03:59.68 Daniel Carvalho Or mediaated by a human because robots today can interact with other robots and program another robot and we see that all the time on the internet. the bots the pop-ups the unfortunate unfortunately the spam and so it's it's a difficult world. But I think the most useful definition it's it's data. What I gave you? um you. 04:20.47 archpodnet You. 04:32.91 archpodnet So So on the second half of that question then how in the context of archeology and your dissertation. What kind of I Guess what kind of robot are you thinking you want to I don't know I don't know if develop is the right word but at least theorize. For archeology in the context of the work you're doing now. 04:54.36 Daniel Carvalho And so in the first year I thought that a simple robot and by that I mean like a mannequin I'm building I'm building a robot named taos. Ah. 05:06.18 archpodnet M. Um, friends. 05:13.76 Daniel Carvalho Of the Greek legend. He's not going to destroy anything I hope but and he is intending to be a humanoid robot a robot about human height with human. 05:32.73 archpodnet So. 05:33.45 Daniel Carvalho Similarities and it exists on the physical world that was my first and it is still my first objective but and then I had a problem because how could it and. Analyze text ah being only a physical form sure I could put some components that enable him to do that. But the solution that I ah encountered was why make a robot. 05:52.50 archpodnet Sure. 06:12.40 Daniel Carvalho To do all the functions when you can make smaller robots let' do a function each one of them so I'm building Talos but I'm building extensions of toows as well. So I'm building drones I'm building. 06:17.74 archpodnet Yeah. 06:31.73 Daniel Carvalho Um, little cameras that I'm I am I am calling them the eyes because they see like us and they are connected to a microcontroller that is like a tiny brain. 06:42.62 archpodnet Tough. Ah. 06:49.20 Daniel Carvalho That has a set of algorithms that say hey this is a very historic artifact and I I say to them good and this this is a roman artifact. Yeah, okay so I'm building smaller robots and. Using them together with the main body and that and surprised me at first. But then I thought okay and this is not very and this is not a. 07:11.40 archpodnet Okay. 07:28.15 Daniel Carvalho Very different from Us. We are archelogies but we have a lot of tasks to do digging analyzing producing reports theorizing. So Why not build something that could do everything. But and at smaller scale and that's how I'm doing my Ph D being building a ah little brick of the wall at the time. 07:57.86 archpodnet We have well that sounds that sounds really cool and and pretty interesting I mean when you were saying Well first off I feel like every roboticist in ah fictional or science fiction circumstance. That says you know I hope it doesn't destroy the world. But. 08:13.87 Daniel Carvalho Sure. 08:17.78 archpodnet The next thing you know we have we have skynet and terminator. So. So anyway, this is this is interesting because I I mean there's so many different I guess. I guess tasks and when you were talking about you know processors and and and having a more human full-sized human-sized robot as your first thought of doing this obviously some of the problems that you run into are some of the things that you mentioned right? like the computing power. Required for something like that to become a human while we're getting closer and closer to having something like that onboard the power requirements and all that stuff is just is just so high right in order to you know, get something to like that to function say out in the field or something in a good way. But I'm just kind of wondering from a theoretical standpoint man with with cellular technology being almost everywhere and you know even even satellite technology getting getting a lot better. We just got starlink satellite for our for our Rv my wife and I live in a recreational vehicle and travel around the country and we got starling satellite in the disc for that is actually pretty small. A lot smaller than I expected it to be and I imagine that technology is just going to be getting smaller and smaller and with those kind of speeds. Are you? So are you thinking about. The robot's computing power and its functional and ah analytical power all being onboard or are you offloading some of that to the internet basically doing some offline processing or using you know banks of servers somewhere else, but the internet to the the internet or cellular networks to communicate with those servers. Um, just wondering if it all has to be on board or is there a local server that these things would communicate with or something something remote that they would communicate with okay. 09:59.33 Daniel Carvalho Okay, and both I think that that the main robot and it exists for that function but another very important function is that is human interaction and human communication. That's why he is an humanoid and be with the power that we have today and the the complex task here is not to build something that is capable of analyzing because that. 10:25.57 archpodnet Prep. 10:38.88 Daniel Carvalho Is almost trivial today. The problem here is to make something that is a true tool and that brings more benefit ah to archeologists. 10:41.59 archpodnet The. 10:55.39 archpodnet Okay. 10:57.85 Daniel Carvalho But in all fields of archaeology. It's very difficult to build a robot that functions on a lab and on the fieldwork at the same time because that involves motion intervals. 10:59.78 archpodnet Is. 11:07.52 archpodnet Um. 11:17.83 Daniel Carvalho Ah, protection from whether it evolves involves something as simple as solar energy. You can't have simply a plug on the middle of the field and. 11:21.31 archpodnet Brett. 11:26.98 archpodnet Food. 11:34.96 Daniel Carvalho The problem here is to make something that is so vast and ib read as the past itself and that is an humongous test and I think it's very difficult to grasp even that and so I'm building. What. We are or what I think that we are more capable now. Um, but 1 thing that we may work towards is virtual reality a lie to all this I think. It will have a major impact on archaeology and I are you really really want to explore that dimension to with my robot I think we can make awesome simulations of the past and. 12:23.96 archpodnet Prep. 12:30.54 Daniel Carvalho Even simulating fieldwork and probabilities and Hypothees and hypotheses. It's It's amazing. The world will we live now. It's full of opportunities in terms of technology and the archaeology can to. 12:42.44 archpodnet Um, yeah. 12:49.46 Daniel Carvalho We need to use it to to our benefit. We can't suggest to close our. Ah yeah, our eyes and think that the past is inaccessible. We have a lot of tools today. 13:05.10 archpodnet Right? Well, you're speaking to the right audience regarding that for sure at least me anyway and I know Paul Paul would love to be in this discussion too. That's why we need to have you back on because I have a feeling he would have a lot of questions for you regarding some of the more technical aspects of this. But. Um, got a few more things to say about this for sure and some some high level questions to ask you moving forward and we'll do that on the other side of the break in segment three back in a minute.