AU Class
AU Class
class - AU

The VDC Gauntlet: Using the Architecture, Engineering & Construction Collection for Virtual Design and Construction

Share this class
Search for keywords in videos, presentation slides and handouts:

Description

Autodesk’s Architecture, Engineering & Construction Collection provides an extraordinary range of capabilities and tools that, when integrated, create incredible insights and make easy work of complex design and project problems. This class will explore the potential and power of digital construction and project management with the AEC Collection. In this class, we will dive into the complex installation of several large tanks in a highly confined facility, utilizing drone technology and reality capture with different software integrations. Some of the software this class will explore is AutoCAD Civil 3D software, Vehicle Tracking, ReCap software, ReCap Photo software, Navisworks software, InfraWorks software, and more. With the integration of all these programs, the project was able to succeed as planned, utilizing found vehicle and building clearances as well as logistical confirmation of the project's ability to succeed. This class will demonstrate the effectiveness of the AEC Collection as well as the virtual design and construction (VDC) methodology.

Key Learnings

  • Understand how a VDC approach can mitigate risk and ensure project success
  • Understand how Autodesk's AEC Collection works together to provide a seamless solution for VDC methods
  • Understand how to migrate data between Autodesk products for different project intents
  • Understand how collaboration at different levels can make an overall difference on a project

Speaker

Video Player is loading.
Current Time 0:00
Duration 0:00
Loaded: 0%
Stream Type LIVE
Remaining Time 0:00
 
1x
  • Chapters
  • descriptions off, selected
  • captions off, selected
      Transcript

      TONY SABAT: All right. Would anybody be upset if I started one minute early?

      AUDIENCE: Nope.

      TONY SABAT: No? All right. I am going to time myself just because I need to make sure I keep on pace. They've been adding stuff as we go through AU. And it's just there's so much stuff I want to touch on and key points that I want to hit.

      So I might go a little over. So feel free to jump in and out as you need. How are we doing? So how's everybody doing this third day of Autodesk University?

      AUDIENCE: Good.

      TONY SABAT: All right? Everybody made it in still? Didn't miss anything from the party yesterday, I hope. Luckily, I didn't have the 8:00 AM session. That would've been a little bit tougher to get some attendance out of it. But hopefully you're in the right spot.

      We'll be talking about the VDC Gauntlet, how we can use the Architecture, Engineering, and Construction Collection to really push virtual design and construction principles. It's more than just design software. Now, it's really expanding into simulation. And many different aspects can be integrated as well.

      My name is Tony Sabat. I'm a VDC/BIM/CAD/drone/reality capture/specialist. I touch a lot of different softwares for SSOE Group. And we're trying to really understand what kind of hardware we can take to integrate into our software and then bounce that back out into the physical. So we're really testing and implementing anything and everything that we can use to really drive innovation in our projects.

      So let's jump into it and see what the actual VDC Gauntlet is. Everybody like the name VDC Gauntlet? It just really evokes that just barrage of software that you're going to be running through. We'll be covering a lot of different stuff today. So I'll try to keep it brief but concise.

      So today, we'll talk about a case study where we're looking at one of our food clients, consumer food brands, that they are doing sugar tank removal and replacement. It required the removal of three different sugar tanks that are about 50 foot tall by 12 foot diameter. We wanted to virtually simulate this so that we could guarantee the success of it as well as make sure it goes through as scheduled. If we can guarantee our schedule, that really is going to help us long term.

      So before we start, the VDC term I think really covers a lot of different people, a lot of different integrations with users. So I'd like to just--

      [DISTANT CHEERING]

      They like it. Can I get a poll from you guys? I want to just-- let's get a show of hands. How many people in here are civil engineers? OK.

      Maybe architects? OK. And then some of the MEP? I don't want to lump you all together, but-- no? OK, a little bit. Maybe structural? OK.

      What else do I have in here? So, yeah, BIM managers, this is kind of a new, evolving role. OK. And then any project managers here? Oh, awesome. Any owners at all? OK, good. Yeah.

      I think this is a great kind of turnout. We have a lot of different people that are affected by these construction principles that we're using. And for clients to understand what these processes can provide to us is really important for getting buy-in as well as working, continuing to work.

      So I'm more of that VDC type role that's kind of like the new BIM manager. BIM manager wasn't really a thing years ago. Because we're trying to figure out what it is. How can we encapsulate what this person does? They're not really an architect, but they're really sustaining the architect. But maybe they are architects, and they're not.

      It's like we have so many blending roles in all that. It's getting really increasingly difficult to kind of quantify what we do. I've seen-- I've met plenty of VDC engineers, VDC coordinators, VDC managers now too. So I think that's kind of the new evolution of these processes.

      So I just want to let everybody know the speaker surveys, if you could fill that out at the end of this, or at the beginning if you want-- we could do it now if we really want to. We'd just skip all that, and if you can just pay attention to me afterwards. But just fill these out if you can so we can understand what you guys liked, or didn't like, or wanted to learn, wanted to incorporate. Or maybe for next year, we can see how we can pivot to get some more ROI out of it and get relevant to what you guys are looking to find.

      Then I've got a little workflow. Here's our first demo of how to do the survey if we really want to do it now. But we'll jump through that. So here's a little bit about me. This is an old photo of me. A little bit different now, but I'm really-- like I said, I really kind of look to try out any kind of technology that we can.

      And with SSOE Group, I'm in the reality capture, drone space as well as pushing all the different point cloud utilizations within our existing softwares too, especially Autodesk software.

      So I really began my training formally in college with an architecture and civil engineering degree, trying to get left and right brain working together. It was a great concept, but it was a lot of work. But with that architecture, we really dove into a lot of the different softwares and tools, like SketchUp and Revit and Photoshop and all these different types of tools you have to learn to present your designs and understand and communicate them effectively.

      So all that software, that software technology, I really leveraged into my civil engineering background. I started with a civil engineering company. And we were using AutoCAD. But it had this weird little letter at the top that was like Civil 3D.

      And I was like, what's that? Why are we doing polylines and all this stuff with this fancy AutoCAD, this Civil 3D? What is that?

      And that's where I started to learn about these dynamically linked objects-- that if you change one thing, it changes all the other things that are affected downstream, just like a lot of our smart Revit tools or SmartPlant 3D or all these other tools too. And it's really that evolution of that technology. So we started to push that and learn that we can start to develop templates and standardize and really streamline our workflows.

      From there, I went to a construction company, working as the kind of the beginnings of a VDC group too. It was very hard that we were either VDC engineers or VDC coordinators. And we kind of all did the software side of things in trying to coordinate all these different models.

      And that was-- I was really the underground portion of that, based off my background. So I was handling ground penetrating radar and utility locating and then tying that all into the model to then tie into a lot of the above ground, below ground integrations as well, whether it be point clouds or whatnot. But from there, I really took a-- I really understood that the technology side of things is where I like to go, where I'm capable in.

      So I wanted to push that into the consultant instructor role. And I began working as a consultant, where we were teaching different softwares, especially Autodesk, Infoworks, Navisworks, Civil 3D, and then reality capture with different laser scanners, actually hardware pushed into the software, pushed into our design softwares, and then into drones and UAVs. I'm a commercial UAV pilot.

      I just actually renewed my license. So it's been over two years now that I've been flying drones professionally. And it's come a long way and really developed a lot. So nowadays at SSOE Group, I'm looking to involve all the disciplines, make sure everything's coordinated, fully coordinated models, and then pushing these workflows with the reality capture and a lot of the artificial intelligence of looking and sorting through all this data for us, computer time versus people time, and then looking through different workflow vetting and implementation, and then any kind of innovation that might show some new innovations for augmented reality or virtual reality or any kind of thing that's starting to hit the market.

      So I think we've had a lot of that this year at AU. And it's really exciting to see how this is going to develop. So let's stop talking about me. Oops, I went a little fast there.

      So the VDC Gauntlet is this case study where we're talking about 50 foot by 12 foot diameter sugar tanks, so huge, huge assets that we need to move. And we've got to keep the plant operational during this project. So this is one thing that the client really wanted, and we wanted to provide that to the client.

      So that meant there was a strategic placement of where we're going to stage this actual project. We had to lift these sugar tanks into place. And I'm purposely kind of hiding it up there for you. Because I don't want to show you all the good stuff at the beginning. But it's up towards the top at the right.

      We had to virtually simulate this to make sure it was actually possible. But as we went and went through it more and more, we realized the best placement for this project is to stage within this almost what turned into be a confined space-- not for you and I, but for the actual tank, the crane, and the truck and everything, fitting all that into this room essentially to maneuver this tank up into place and up and over all the obstacles in the way.

      We realized that we had inefficient, ineffective as-built data. This is an incredibly complex campus with very dynamic grade changes throughout. And so reality capture was going to be the only way to really solve that problem and paint the picture that we needed to design off of.

      And then we'll talk about this collaborative campus approach that we really target with clients that have large facilities. Because it really helps with downstream data and master planning. OK, that's good there. All right.

      So one of the pillars we're going to be talking about is going to be collaboration, this campus approach, understanding your company's capabilities, and then identifying your subject matter experts. And then we'll talk about reality capture, really setting your project's foundation, and then drones and UAVs as another subset of reality capture. But I do want to specify and kind of spotlight the drones and how they were very, I would say, pivotal for us in this project.

      And then of course the simulation and actually simulating and modeling to our needs, I'm going to say that modeling to our needs versus full simulation. Because VDC can be deemed very, very detailed and very elaborate in all these different tools and techniques that we can use. But it truly isn't. It's really cost-effective when we're integrating so much software at the end of the day when we can already utilize the tools we have.

      So first one is collaboration. Is there such thing as too much collaboration? Probably, but again collaborating to your needs. We'll say that again. But the VDC Gauntlet project that we're talking about today really was born out of another project. And that was this campus connection renovation projects, where we're connecting the central building to three exterior buildings.

      And this is where the dawn of that reality capture data that we used downstream for our project really began. It was these covered walkways that they want to connect the interior building to the exterior. They have-- it's a food plant that we need to keep in very clean, sanitary environments. And jumping in and out of buildings can delay a lot of workflows, whether it's employees moving in and out or moving product through it as well.

      So a controlled environment, incredibly complex campus, this dynamic grade change I was talking about. From the bottom of your screen is about 20 foot higher than at the top of your screen. You can see kind of some of the truck path going through. And that's about 20 foot difference.

      So reality capture was the only solution that would really help us fully be able to coordinate the design for these connectors. And then we kind of grandfathered or moved that data progressively into our sugar tank process. So this is where the connectors project is.

      We utilized high definition laser scanning in combination with UAV photogrammetry. We registered them on the same target system so that they would come in right as they needed to together. We put them on more of an arbitrary coordinate system at 2,000 2,000 so that it would talk well with Revit. Revit it is going to be the main user for this data when we're touting the connection design.

      Not really relevant to our sugar tanks, but I want to point out the high definition laser scanning in combo with the photogrammetry. We used photogrammetry to really cover the gaps that the high def scanning wouldn't get. And then that would give us kind of a whole idea of the actual picture.

      This is where our sugar tank project is, way up top right over there, that little guy. And it's about the same size where we're actually going to maneuver the project in and out of. So that's really going to-- that's why I'm calling it the Gauntlet. Because it's going to berate us with all the different softwares we have to integrate and utilize.

      The reality capture we added on the secondary part of it. Because it was kind of like-- it almost felt like a scope creep idea or a process. But as we understood the needs that they needed to do, we understood, yes, using photogrammetry would be a great addition to the project for you guys. And it wouldn't actually impact the scope. And we'll talk about the differences of the data in a little bit.

      But we wanted to understand what their needs were. It was kind of like a simple ask and a simple understanding of, what are our capabilities? Who can I ask for if this is actually going to fit my needs? And then how would that look in the scope of things in the construction schedule and in the costs?

      So I want to kind of tie-in with what the VDC things that we kind of glossed over in the beginning-- was kind of this collaborative approach, and just basic understanding of questions that we need to ask for different projects, and over scanning that-- understanding that we have these projects for clients. We have multiple projects going together for one client. Can this data be used downstream? Can we over scan a little bit, over capture a little bit to use that for our purposes? And understanding, then, your company's capabilities, right? Is this something that we actually are doing? Do we have this VDC tool belt that we can use?

      It's kind of like, remember the beginning of when BIM started to come out, and people just knew about it? They didn't know what it was. They don't know how to ask for it. It's like, can we have a BIM on our project? Can we do BIM for this one here? Same thing for VDC. It's like, can we do VDC here? You know, you don't really know how to ask that for some people. And so it's great to communicate that and give some more learning and understanding topics for people to ask.

      So we all know that Virtual Design and Construction is more of a project delivery method. It's an overall umbrella where we're able to capture and simulation all these. And BIM, there are all these, like, subsets of it. The earlier we can integrate VDC, the better. We kind of came in a little bit later for the whole project for the Sugar Tank replacement one. But we were still able to keep it effective before they actually began the construction schedule.

      The next point is just identifying your subject matter experts and knowing who to ask about integrating reality capture within your company. Are we doing drones? You know, what kind of efforts are we trying to push? And then, who can I address my questions to, to assess the needs of the project and assess the capabilities we currently have?

      And then that Holistic Campus Approach, I really want to talk to. With some of the larger clients, they have just campuses. They don't have just one building. And this over scanning, this over capturing can really be effective from project to project. And the client will definitely like that as well for them having an entire campus mapped and looking at different master planning capabilities and streamlining and scheduling out products in coordination with [INAUDIBLE] together.

      All right. So that's our VDC portion there now. Excuse me. I wanted to touch on those. Because it can be easily glossed over, where people won't even-- project managers might not even think of a VDC approach or utilizing drones or utilizing reality capture. For a small project where we had to just design the tanks kind of in space in relation to existing drawings, it wasn't really thought of to do a VDC approach. But as we developed this complex confined space that we're going to have to operate within, it seemed all that we really need to look at the detail of how this is going to physically work.

      And the basis of that is going to be Reality Capture, right? Reality capture is now the foundation-- oop. Here we go. Reality capture is becoming the foundation from which we design off of, from which we simulate off of. And the more accurate the reality capture data that we have, the more accurate in our simulations we'll be downstream, all right? The better our foundation, the better our design will be.

      And we really think of reality capture as a subset of VDC, giving us the best tools to then design our tool belt off of. So this is a great image. I wanted to use this one. Because this is one of the buildings on the campus.

      It was snowing that day. So we even captured the snow. But we utilized high definition laser scanning in conjunction with UAV photogrammetry to get that holistic approach over the holistic capture.

      But you can really see the detail. We're using high definition laser scanning. Let me try this guy. And you can start to see the deviations in the rocks with how accurate this is. And this is a unique design. This is for the connectors, basically, initially. But the high definition laser scanning really helped out downstream for our product as well.

      And it's really difficult to figure out where we're going to tie-in. Are we going to step back from these rocks here? We had an architect at the office that was just pulling his hair out because of the way we're going to be defacing this building from these connection projects. But it's really important to understand the capabilities of what you need for the project, and then what kind of tool is going to fix that-- that solution for you.

      So I've got here an image from the high definition laser scanning where we're looking to get millimeter accuracy at 300 foot range, at the minimum, OK? That's different than photogrammetry. But this is essential for us to get that project foundation. And we're actually getting the-- if you look all the way to the right, we've got these wires are even being mapped, so we know that when we're tying into different utilities or when we're routing these connections over these existing lines, we're not actually going to interfere with the actual lines or the whole utility.

      So again, though, with High Definition Laser Scanning, the more accurate it is, the more the cost is going to increase, right? And the scalability and the maneuverability of it is going to go down a little bit as well. Because you want it to be accurate, and you want it to be right, OK? When it has to be right, millimeter accuracy can't go wrong.

      But then drones and UAVs, now, come into play. There's some trade-offs, right? We're not going to have millimeter accuracy from a drone. We're going to have less accuracy, we'll say. We'll just say that to keep us a safe statement there.

      But I do want to address drones, individually. Because there are so many capabilities coming out with drones, like, almost weekly with what we can do with them. And I think the lowest hanging fruit is really Asset Inspections.

      A lot of the drone industry gets clouded with a lot of the cool videography and cinematography and all the vloggers and everything that, really, asset inspections-- video and photo, GPS located photos-- can really help you with understanding your assets that are existing and kind of addressing the workflow we have to do for them. But with more and more sensors being integrated on them, you can really start to get close up to some of these assets, these walls, you know.

      They're essentially getting into, like, Confined Space inspections with how tight these drones can maneuver autonomously or remotely. Some of them-- I don't know if they're here in the Expo-- but some of them actually have their own, like, hamster cage around the drone, so that you can actually run into a wall, and just drive right up to it. And then you integrate lights and all these other features. And you can really start to generate some value off of them.

      Photogrammetry is what we'll talk about for drones, specifically. But photogrammetry wasn't made possible by drones. It was made popular by drones. Photogrammetry has been around for, I think, since the 60s or 70s with basic plane photos and overlaying those.

      But has anybody ever used Autodesk Memento? Remember Autodesk Memento at all? Yeah. That was, like, Autodesk's first iteration of 3D model photogrammetry. If we want to capture that projector or something, you're going to take photos of it. And they're going to stitch together your 3D model.

      And then Thermal Inspections are along the same line of sensors. The more sensors we can put on these things, the better. And it just seems like, natural gas sensors, thermal sensors, there's so much to add onto it.

      But LIDAR. LIDAR is very exciting. Because it's getting rid of that photogrammetry aspect and trying to see how we can start to create these autonomous LIDAR laser scanning machines. For any reality capture, I would say, enthusiasts, has anybody seen the movie, Prometheus, the predecessor to Alien? Kind of the holy grail in reality capture is in that movie where they're trying to map this underground region and trying to understand-- they're traveling back to figure out where their human race evolved from and all that stuff.

      But they want to understand the conditions, so they can figure out where these aliens are hiding. The guy takes out, like, a softball and throws it up in the air. And it, essentially, starts mapping and LIDAR. It's shooting a laser scan as it's flying and hovering itself. And then it starts to move autonomously through the whole space.

      So I mean, you've got drones there. You've got laser scanning there. You've got autonomous flight there. But at the same time, it's also beaming that data back to the mothership that they're capturing. And it's live streaming the data as it's going. So it's like a 5G connection where you're streaming point cloud data. And that's kind of like the holy grail of us reality capture nerds out there.

      But it's really cool. Because a lot of that technology is downstairs. And they're getting closer and closer to at least getting usable data, competitive data that will rival some of our terrestrial scanning capabilities as well. So that's exciting to see in the next couple of years.

      So this is the site that we captured with the photogrammetry of the drone. As you can tell, it was snowing that day. I'm from Michigan. So if you don't like the weather in Michigan, just wait 10 minutes. And it'll change for you.

      So it was snowing 10 minutes on, 10 minutes off, 10 minutes on, 10 minutes off. And it wasn't the most ideal conditions to capture with a drone or a laser scanner. But we capture it as is. You know, we took a photo exactly where it was.

      So it can really add a little bit of complexity when you're starting to look at deviations in photos and everything is almost the same color, right? So the snow really threw in a couple of wrenches in there. But it's OK.

      It was actually good to capture in the snow. Because when we were talking about this with some of the construction managers, we were showing it. And they were like, oh, we can really see the actual truck pass and how these logistics are moving through the campus. So that was a little side benefit there.

      But I want to tell it the size of this. Because the maneuverability of these UAVs is what's really making them effective and powerful. This was about, I'm going to say, 50 acres that we captured within a couple hours.

      And why we were able to use UAVs and get some actual data off of them was because to increase the scale of the capture, to over capture with the drone, for our project, the Sugar Tank project, it took an additional 10 minutes to get the photos we needed to better stitch together the photos we needed to get that point cloud to get that added accuracy. So it allowed us to scale and incorporate into many other projects incredibly easy.

      So this is the Google Earth image of it. It wasn't snowing that day. But we can tell, each of these blue dots is where we took a photo to better capture the site.

      The purpose of the drones is that a lot of them are GPS tagging their photos. So it's better to orient the relationship of them to each other, versus some of the 3D model photogrammetry where you really have to make sure you have the proper overlap or else you're not going to be able to get the right model out of it. You're going to be missing a lot of gaps in your data.

      We did two different flights for this mission. We did more of the mapping, where it's kind of like a slalom pattern, if anybody's familiar with drone capture. Anybody fly drones here, or use drones on a project? A little bit-- couple people. OK.

      So the mapping is kind of the first type of flight where you're going to kind of set the foundation for the photogrammetry model and kind of standardize your levels and your 3D elevations. After that, then, you're going to use more orthogonal photos to capture the facades. Because we're looking to capture the building side to side, as well as for the Sugar Tank project. We have to capture the integrities and the differences between the different buildings within that area.

      There's a trained canopy in there. There's a conveyor. There's a conveyor structures within there. And they've also got the tanks and everything. So a lot of photographs to use to model this.

      And with the photogrammetry mapping mission, it was really good to begin with that to set our standard. That's mainly, like, where photogrammetry began, the preliminary mapping stages of it. When it was talking about Memento, Memento was more of just 3D objects floating in space type of thing, where you have to apply a scale, potentially, to it as well.

      It's nice to do that with the drone here. Because you can do-- the GPS scale can be applied to that already. And you can count on your relative accuracy. With the mapping, it was kind of, like-- how do I want to say this? You have to set your specific amount of overlap.

      So you have a side lap and a front lap that you're making sure that you can properly capture that. And the more photos you use to capture it, the more accurate, the more dense your data is going to be.

      Same with the orthogonal photos. Really, you can just start to do it on, like, a timed mission-- you know, a time delay-- every three seconds, you start taking a photo. And then you can kind of map it yourself and just making sure you're looking at the right object as you're capturing and the GPS tags are being recorded along with that.

      So now the difference. So this is high definition laser scanning versus UAV photogrammetry. They are two different sensors, two different capture methods. And they do have differences, right? They are all relatively accurate. They're not all as accurate as some of them claim to be.

      I was actually going to be in another presentation talking about the differences of accuracies in reality capture tools and softwares. But they scheduled it at the same time as this one. So I'm going to give you guys my rant now.

      High definition laser scanning-- we've got an actual return, right? We're actually hitting a wall, coming back, and measuring that. Photogrammetry is different. Because it's more of a-- it's a measurement between capture techniques, right?

      So have you guys ever heard of the structure for motion technique, where it's, essentially, if I take a photo of this building, and then take another photo with maybe 50% overlap, based off of the difference in position of where that camera was, can then tell me what that building elevation is. That change in pixel then determines that this building is 30 feet, something like that, OK?

      So it's more of not like a single return. It's multiple returns based off of each other. So overlap is very essential for photogrammetry. But photogrammetry can get a little bit, I like to say, like lava-- looks kind of like lava, right? If we're trying to capture this room, these hard corners, a laser scanner is going to get. It's going to be right angles. And you're going to know it.

      Even with some of the lower grade ones with, like, the Leica BLK360 type stuff, you're going to see those sharp corners. With photogrammetry, it's going to be difficult. Because it can tell the difference in pixels. But it can't tell that that is actually a corner there.

      And the amount of photos that you take is going to help. But it's still going to kind of roll. Your joints are going to roll. And you're going to have that-- it's going to look like your building is kind of melted to the ground a little bit. So it's going to be a little bit different for the photogrammetry. And you just have to understand that when you're doing that. And if you're trying to get dimensions off of building corners photogrammetry, just be cognizant of that and understand where you're snapping to, and where you're trying to get your measurements from.

      This is the comparison of the two. This dark black line there is the laser scanning. And the photogrammetry is the kind of fuzz around it. But you can see, a majority of the measurements are where they need to be, right? So it's relatively accurate. It's going to do what you need it to do. But you do need to know where you're snapping to and where you're kind of extracting your methods from.

      Because if you snap to a different area of it, you might be six inches off, four inches high or four inches lower. It's going to vary. So just something to keep in mind as we go through. I'm going to jump into my first demo.

      I'm a first time speaker at Autodesk University. So I appreciate you guys being nice to me. But I was also told, don't do any live demos. So I decided to do four live demos. And we're going to see how this works.

      So I just want to give some kind of tips and tricks. We've got a lot of people in the area, a lot of people from different backgrounds. And hopefully, everybody can get one different thing out of this at least.

      So this is ReCap, Autodesk point cloud engine. Basic zooming in and out. If you're going to orbit, you're going to have to hold down the right mouse key and click in there and orbit over that. This is all the data together.

      A lot of people like the cloud to cloud registration of different reality capture techniques. We like targets. And we like to be 100% confident in where these are located. So we use the same targets for a high definition laser scanning as we did with photogrammetry. So everything just drops right in, right? If it can just drop right in, and I don't have to worry about it, that's what I want. I don't want to be piecing it together and interpreting it, OK?

      I want that to be done in the field. So if I turn off-- right down here is our little Layer Manager. We've got the photogrammetry here. I can turn that on and off. And you can see the difference, right? Laser scanning, we're not going to put it on the roofs. It's not really necessary to that connector's project, right? The connector's project was the initial effort that utilized the high def laser scanning data.

      The photogrammetry was more used for our Sugar Tank purposes. And so that really was able to get a lot more data that we needed for this area over here,

      If you guys can see, there's obviously different colors of snow on the ground. That's not true. But we used two different scanners when they were capturing it. So one is a little bit older-- different color rendering. The other one's a little bit newer. So that's why you've got differences in the color there.

      But if we zoom into-- like, this is where that one slide was at. And if we zoom in on this data, you can see it's incredibly dense, incredibly detailed. You're getting these jagged edges of the rock. You know exactly where everything is going to be. Millimeter accuracy, all right?

      If we turn on a photogrammetry now, it's going to be less accurate. But it's going to be relatively accurate. Like if you look at this area here, everything's white that is trying to capture. And it actually did a pretty decent job.

      But you can start to see noise around there, kind of spray. So you have to understand if you're going to be interacting off of this for some measurements or for some design, where you're actually snapping to and be cognizant of that as well.

      All right. So let's look at the data. When I'm digging in the data, and I'm trying to understand how accurate it is or where it's deviating or how it's behaving, I like to cut really, really small sections of it, right? I like to get some sample sets where photogrammetry other techniques like that kind of get, I don't want to say dis-- almost like disillusioned is that they claim 2/10 of a foot accuracy using ground control points, using this using that. And you're not wrong, right? But where you're measuring from is usually where you're exactly tying in and telling it in the photos what this elevation is. So, Yeah. Hopefully you're going to get 2/10 of a foot accuracy. But as you start to cut sections of it-- this is the limit box within ReCap, almost like a crop state. And if I can hover over this, different planes will highlight. And I can start to click and drag into it. And I can start to cut a section through it.

      And I can go less than a meter. And I'll hit Confirm. And then I can go to the right. We'll go to our front there. And we'll zoom, Zoom way in there. And you can tell that there's already deviations in the photographic tree data. If I turn it on and off, can you guys see that, OK?

      OK. There's our laser scanning. There's our photogrammetry. It is relatively accurate, very, good. But if you're snapping at different points, if you're looking at this point, and you're four inches above, if you're looking at this one, you're six inches below. And this term, accuracy, is really not very accurate when you're trying to describe the accuracy. Because in different areas, you're going to be positive plus or minus that 2/10 of a foot accuracy.

      Again, relatively accurate. But that's why it's kind of-- I always try to tell to people that it's all an education process and understanding what kind of data you're getting, And then where you're going to be snapping to extract off of that. If you're going to be designing off of that, are you snapping to where you need to snap to? If you're measuring off of that, are you measuring where you're measuring to? Because it's all about where you're picking, right? A lot of this laser scan data is you actually picking that point to then say, this is what elevation it needs to be, et cetera.

      So quick little demo there. Limit box, layer manager, turning things on and off, and then we can reset that limit box. Hit confirm. And there we go. So there's our first demo. Didn't do so bad. Still live. All right. So that was ReCap and more of our reality capture. And we'll regroup.

      All right, so next up is the virtual design and construction, more of the simulation needs of things, right? Actually simulating this and how this tank is going to move, how this crane can pick it, where they can all fit and proportion themselves, and how this tank can all go in there. So reality capture really set a great foundation for us to work off of, and now we have to figure out how these existing conditions can work. You know, we're going to design it virtually so that we can design it in reality.

      As you can see, the image on the right is on the right. That is definitely a confined space. The guy that's taking the photo was actually underneath the actual conveyors. The crane is right up to the tank, tank looks like it's five feet from that train canopy right there. It's going to be a logistical nightmare to actually do this, so we have to think about how we're going to actually simulate this and prove that it can be done, and prove where you need to be placed as you're going through it.

      So again, keeping the plant operational, this confined space method, and then the crane size and placement. And then as well as the truck and trailer logistics, and we'll talk about that too, because you've got to get the tank to the site. And it's not your average tank. It's not your average truck and trailer that's going to get there. And then how we can get that and integrate it into the actual process.

      So confined space simulation, this is the first, right? We're setting our groundwork. Let me get my laser pointer. Our playground is like right here, right? Almost like the size of this room, I'd say. And so there's several different stages we're going to have to investigate, but we have a lot of things we've got to figure out. I've got dueling clickers. So we have to first put in a crane, right?

      Where's that crane got to go to lift that tank into place? And where's that-- come on. Come on. I got some really great clip art for you guys. Where's this tank going to go? And then-- there it goes. Where is that semi truck going to go to back that up into there so that you can lift that off of it, and then move it around, navigate it around, pick and place? So it's going to be a little bit to make sure that this is feasible in reality.

      The existing train canopy right there is kind of the wrench in our gears here that's limiting us, and we want to eliminate as much waste as we can. So let's start off with our basics, our sugar tank modeling. We're going to crawl before we walk, OK? To model a sugar tank, you're going to draw a 12 foot diameter and you're going to hit extrude up to 50 feet. And there you go, you got your sugar tank, right? This VDC stuff's easy, you know? Anybody can do this, all right?

      So that's our foundation. That's where we're starting. Next is, how can we maneuver this, OK? This is where I want to start modeling to your needs. A lot of people love Navisworks simulate, and a lot of it is this key frame animation, where you can animate a tank actually moving up, being placed in place. You know, it looks awesome. Very cool. You can spin it around in 3D as it's moving. You get all this great video off of it and everything, but is that really what we need right now?

      For us to do all that, that's going to take a little bit of time, and this is a small project. We got to keep our budget on track. So the rotation, the actual tank lift is going to be a pretty complex part. It's not just a simple rotation, right? If you're going to do a vertical structure, you're not going to pick it at one end and rotate it up. We've got that train canopy that's really hindering us. So it's going to be not only a rotation but a translation, a movement.

      So you're going to rotate it and you're going to translate it. So it's more of a compound movement, and each axis of that tank has to be under control. So the crane is going to have the one end, there's going to be a high-low at the other end. So they'll both lift it and support it up, truck will move out of the area to give you a little bit more ground clearance, so then rotate that tank up and over. And then maintaining proper clearance from the train canopy so that it doesn't damage the asset during installation.

      So this is how we do this how we went about it we just modeled each stage of it from beginning to end. White is kind of-- I did like a gradient to show the actual stages of it. Beginning is white, end is in gray. And we just use it all as one mass, right? Because nothing is going to be in that area during this lift. Nothing is going to be interrupting that path, so we can assume that we can model this as one mass and just overlay this exact thing into the model.

      We also modeled this yellow bar there. It's kind of the train canopy end point, and I offset six inches from each just to give it a safe buffer so that we would know that we can make this rotation without scraping or scratching anything. And then just in pink is the different rigging material. But very, very rudimentary. We could've done way more. We can always overdo things, but we have to understand that kind of simulating to your needs.

      So now I will go on to our next demo with Navisworks. And I didn't even do four of the same program simulations. I did different programs for live demos as well. So Navisworks is-- this is all the reality capture data in Navis, exact same stuff we had in ReCap. Same zooming in and out functionality, but you're going to hold Shift and the mouse wheel to orbit and you get into the detail area there.

      This is our site, OK? This is where we have to maneuver all that. So that's going to be fun, right, to figure that out. So with Navisworks your selection tree is almost like your project properties in understanding what's in your model. Luckily I've already got the tank mock up in there. And this is just from SketchUp, too. So we just integrated that into it and imported it.

      And Navisworks is a great program because it can really consume a lot of different data types. But this was very interesting to look at and understand. You can see we've used kind that fixed-- the starting cylinder is kind of like our baseline of where we're going to pivot our tank up off of. But to get a better idea of this, we really need to run sections through it. So let's show how to do a section real quick.

      We'll go to our viewpoint tab and we'll do enable sectioning. And we can zoom out there, and we've got different planes up there, kind of like ReCap's limit box. And you can see where we placed that as our fixed point and give us plenty of room to simulate that. With Navisworks, the way you do almost the limit boxes, you can do planes, you can do bounding boxes.

      And when you do these different planes you can turn them on and off, and then you can even link them as well. So when I start to move this, I want to get more sections through this. I can start to extract this here and I can really get under there and see how this is hitting the train canopy and how that's going to interact. So now I have to figure out how much space is left to then put a crane in there, OK? That's the fun part.

      So talking back and forth with the logistics crew that's going to be actually living this into place, they wanted to know, well, if you can do this, then what's that distance from the edge of the train canopy to kind of the extent of that tank rotation? So then I can figure out where I need to put my crane. So, well, we can definitely do that. So we'll go to the Review tab and we can start to extract some measurements from Navisworks.

      And so I know the extent of the tank is right about here. And you can see they'll snap to that edge there. And, well, we kind of want to get it right exact. It's kind of tough to figure out what I'm snapping to. The good thing about Navisworks and ReCap as well is that I don't really have to snap to where, you know, maximizing kind of exactly where I'm at with that plane, that train canopy plane.

      So I can actually click right here and you can see I'm getting deviations now on each axis. And I know exactly that this is actually 32 foot offset from the that train canopy edge. So I don't need to actually do the ortho or simulate them in just specific coordinates or specific axes. So now I can deliver that and say it's going to be about 32 feet you're going to have place off of that. And we actually use these in our project documents, relaying this in court and communicating back and forth with the project teams.

      And that was our Navisworks demo, so let's go back here. There we go. So we'll use this same method for other measurements, too. The horizontal distance from the train canopy to the first conveyor structure, right? This is ReCap. This is some of the high definition laser scanning data that we were able to still get. And then the horizontal distance from the edge of the building to the tank from the Navisworks data and the photogrammetry set.

      Very helpful, very useful, as we're progressing through the project to communicate dimensions and tolerances that you need to maintain clear of. OK, and then here's the actual tank lift simulation, just again to talk about the actual extents of that tank from the train canopy. And then now I have x amount of feet to place my crane and to understand the differences there. So that was pretty easy part, right? It wasn't too bad for VDC simulation, right? It's not too shabby to learn.

      Next is the crane, though. Crane is going to be easy too. I've got this clip art, it looks really easy to start simulating. But actually it's hard to get into it. It's like, oh, my goodness, there's so many specs that they have for these things to make sure that they can lift what they can lift and fix their orientation and their axes and the x and the y of each of their overall width.

      The crane arm, the boom arm, there's so much stuff to understand in this. To model this would probably be an inefficient use of our time. There would be a couple days worth for a good modeler to start modeling this. So the way we simulated this was we started searching the 3D model warehouses everywhere. And if anybody's familiar with SketchUp Warehouse, they've got a lot of great stuff in there, a lot of free stuff you can use. And they actually had the GMK5180 crane that we were going to simulate with.

      Lucky for us, right? But this is a free model library, and I just want to point out the disclaimer that all these models, if they're free, you know, you have to really QA, QC them to make sure that they are as accurate as they claim, or that they are represented as they are. But it was great to get this, but there's a lot of different 3D model libraries we can start to get different data off of, from AutoCAD, Revit, MEP. All these different softwares have their own 3D models that we can start to utilize and then extract to our own uses, OK?

      But I want to do a quick demo of this to jump through why it was important SketchUp was so good to be used for this. There's a SketchUp model that it had. Let me clear my dimension. And luckily I have the GM1K right here, and we're just going to unhide it. And we can show how we actually placed it now. I wanted to show this because this isn't what it looked like in the last frame, right?

      It's usually, in different iterations, it's clamp down. There's no boom arm. What we actually could do with this model is, why Navisworks is so great is because you can start to drill into this model and look at the different assemblies and groups within it. So right here I have this blue is selected. This is my kind of like my plumb bob to know exactly where we're picking that tank. And then this is the actual crane model.

      And we can drill into that a little more, and we have all these individual measures. So these are just individual 3D objects, not necessarily groups. And then we can start to go through these different instances. Navisworks is recognizing these are native in SketchUp and it's reading them. And I can see I didn't actually simulate it correctly without the counterweights on there, so this thing would be tipping over in real life. But I can select this instance now, and this is my actual crane that's going to be rotating and picking this in place.

      And so it's nice to be able to filter this out. And now I can go to my item tools and I can individually either move or rotate this arm into place. And if I hover over this plane here, you know, that's not the right rotation axis. I can actually move that rotation axis and click and drag it around where I want it to snap to. And it's going to snap onto that object that I have selected.

      And we can position that and I can isolate it in the green and red. And now I can zoom out and I can rotate this as it would be rotating on the actual crane itself. And I can fix it around that axis. So really important to understand where the crane is going to be parked and then how the arm's going to move and simulate. So Navisworks can read all that and help us dig into the details a little bit more. OK?

      So there we go. We've got our tank, we've got our crane. Next up is actually getting it to the site with this different truck simulation now. This is all we were given with the truck. We were just told it's a double drop deck trailer that's going to be used. It's a unique scenario, too, because it's a 12 foot diameter sugar tank. Highway transportation limits to 14 feet.

      That doesn't give you very much room to actually lift this thing up. So that's going to give you maybe two feet of the trailer and then your ground clearance. And with the amount of spread that we had to do, their actual project trailer length turned out to be 110 feet long. So this is quite big, almost double the average truck and trailer length. And to simulate this, we needed none other but Vehicle Tracking. Has anybody used Vehicle Tracking before?

      Oh, good. Vehicle Tracking I'd like to tout a little bit more, because people think it's only for civil engineers because it's only solely in Civil 3D. But a lot of people can use it in AutoCAD as well, and I'd like to push that to architects who were trying to do different simulations for, you know, logistics moving around a site. Can I move this type of trailer in and out too? And you can. Again, it has its own 3D model library that you can start to use.

      And it did not have exactly what we needed, right? So we had to again tweak it and customize it. One thing we had to do was get the length right. First off is the length. The trailer length about was, I think was about 83 feet total plus the truck length itself. And then an average deck is about eight foot wide. We had to simulate it to 12 feet because of the actual tank itself.

      So we have to make sure that-- we just actually stretched out the barrier of the actual trailer length to 12 feet to make sure if we're driving past buildings or fences we're going to actually hit our clearances that we need. So this is the first simulation we did on the site. Our role was to make sure it can maneuver around the site. It's going to have to backup underneath that train canopy.

      I mean we used a couple of different softwares here. This is Civil 3D, but we geolocated it so that we could actually get this Google Maps area. And we had to make sure we were avoiding the white area there, which is a lot of the trailer staging, and then we're entering from the south. So this is our assumed conditions we have to figure out, right? So entering from the south, avoid that, don't hit that, can you do that? And we said yes, yeah, we can do all that.

      This is Civil 3D. You can to do all this in AutoCAD as well. Geolocation is actually a native AutoCAD functionality, not just a Civil. But we had some extra time on our hands, so we simulated it going into the site. This isn't any fancy area. This is a normal neighborhood that hasn't taken into account truck and trailer lengths for special access.

      So we simulated it entering from the south again and we found that it was hitting a lot of places. If I can get my clicker to go, there we go. It was hitting-- and we had to over steer onto the site. We were hitting utility poles on both sides. We're hitting the fence posts. With our 12 foot length, we were scraping the building because we can't turn into there fast enough, and then we're also hitting the fence entering in to the site.

      So we pushed this back to the actual logistics crew shipping the tank on site and, you know, they trust their methods. And you don't you want to say, you're wrong, but you want to say, hey, can you double check that? We're simulating this and we think you might have some issues. And they ran out and they actually have their own kind of like, an actual line that they run to represent each different axis as the trailer is moving throughout the site. And sure enough, they found the exact same thing.

      So this was one of those opportunities where it's like a value added opportunity where it's like, had you not done it this way, what would have happened? You know, what kind of schedule would you have impacted there? What kind of damage could you have potentially hit on the asset itself? And it's really hard to quantify these tips, these moments, as well as you're going through and simulating these things.

      So we advised a rear articulated trailer with vehicle tracking, where you have an articulation between the truck and trailer and then you have an articulation at the back of the trailer as well. So there's actually the guy at the end of it that can remote control the trailer independent of the actual truck and trailer's coordination. And we proved that can be done.

      You're only going over steer onto the lawn of one of the neighbors, so you can handle that pretty easier than actually taking down fences and all that stuff too. So we're pretty much done, right? That's the whole project? Not quite, right? Like I said, this is 110 foot trailer length with about six inches of clearance from the bottom of the trailer to the ground. Something's going to scrape, right?

      So we had to figure out how we can vertically analyze this, too. So we did a vertical clearance simulation with all the steep grades on the site we really need to account for that. But this is all the photogrammetry data that we were using. This is pretty much all the elevation data that we had. And this data is great, but over on the right there, where the parking lots are, where they're going to be actually driving through, isn't the most accurate. We didn't capture that with the drone.

      So how can we address that and how can we get accurate elevation data? Well we used Infraworks. Anybody used Infraworks before? No? Its one of my favorite programs to use. I could talk about Infraworks all day, but we're going to focus just on Model Builder. What Model Builder does is a feature that kind of paints your existing conditions for you, very preliminary.

      It takes a lot of existing USGS data, elevation data, combined with satellite imagery, overlays, all this data. Gives you roadway data, it gives you building data, potential building heights, too. This is the site. None of this was modeled by us. We just windowed our area, we use Bing Maps, and we said, yep, do this area. And it did it all for us. And so that gave us our existing conditions.

      I think I'm running low on time, so I'm not going to-- I got to skip my Infraworks demo, so I'm only going to do three. But the vertical clearance analysis we then push to Civil 3D. So Civil 3D and Infraworks can talk very well together. Infraworks has an actual Civil 3D export directly to it. So you export an IMX into Civil, and this terrain data then can be used as an actual Civil 3D surface, which Vehicle Tracking can then understand and simulate off.

      And so as I said earlier, the truck can raise and lower four inches. Anything above that is really going to deem a reroute. And when we ran the simulation we for sure hit that you could see in that red area, red and pink area, that's where we were initially planning this, right by the guard shack. We were hitting from five and a half inches below grid with this current truck and trailer combo that we had.

      So here again is another value added where we said, I mean, if you do that, you're going to damage the trailer. You're going to damage the asset, potentially. What if you entered in from the east, OK? And we push this back. And sure enough, this is one of those value added scenarios where it's like, had you not done it this way, what would have happened? What kind of time delay would you have gotten and what kind of cost impact?

      And then to communicate all this we export it from Infraworks or Civil 3D into Google Earth, lightweight file format a lot of people can use. It's free. And we can communicate this and show hey, this is where we're seeing. And you can even start to geolocate yourself too in Google Earth, as well. So it's a very quick, mobile collaboration tool that we could use back and forth. So that's how we fully proved everything.

      And this is a really cool project that we've got to be involved in. Because we were we were there from the beginning for a different project, over captured, designed everything, simulated it, and then we even got to record it actually happening with the drone. So I got a video for you guys that we can run through.

      There's that crane and that's the actual-- you know, that little site right there is what we need to do all this for. Here's almost the plan view. So that was reality, this is virtual, there's Google Earth, saying just how we were planning it. And then you can see here that he's actually parking under the conveyor structure there. With some of us laydowns he really had to scoot it in some more.

      So sometimes that tank will actually go underneath the actual straight down from that, then, the extent of the boom arm. Google Earth again. You see the vehicle tracking path going inside it with the cylinders right there on the other side of that canopy. And this is a great shot of the truck driver moving independent of the trailer and the articulated rear end as well.

      See the guy in the back there? He's just got a remote control, he's tweaking that left, right, left, right so that thing can line right up as you need to. So now that the tank is in place, you have to lift it off. So this is sped up 20 times as the actual speed of it. This is actually like a two hour, one and a half, two hour process of each lift.

      And this is kind of where the communication from the actual crane and the high-low are talking back and forth, and that's where you get that shimmy dance where it's up, over, up, over, up, over, up, over. And here's where we simulated it in Navisworks. So it probably was quicker for us to simulate it than it was to actually do it, but the actual communication was the huge part that they needed to really, really depend on each other for.

      And because it was all be able to be done within almost this confined space, you're going to see truck traffic be moving around the actual site during the lift, and maintaining plant operations during the whole process. So again, it was about a day per tank for lifting and for staging everything and getting the truck and trailer into the area. And then they just set it up into place, dropped it into place.

      So it's like we're almost at the end of our time. Just to summarize, the removal and replacement of these existing sugar tanks, using that clamp collaborative approach and over scanning, over capturing to use this data downstream, using reality capture and drone technology. Modeling all our sugar tanks, our cranes, our vehicle, our truck and trailers. And then identifying those value added scenarios. Had you not done it this way, what would have happened?

      We really try to look to quantify these results, But Sometimes it can be difficult to communicate that back to the client to say, we're going to return our fee to you guys if we use our techniques. So does everybody know why I called it the VDC gauntlet? It was pretty apparent, the kind of barrage of software we were hitting it through. But is anybody an Avengers fan at all?

      Because actually I was looking at the Infinity Gauntlet they were talking about in Avengers where you have all six of these Infinity Stones, each one representing a different super power, whatever. When they're all combined into one, you know, you've got the power of the universe in your hands. And we had, you know, six different softwares here but they only have like three different colors.

      So it's like, well, you could use Inventor to model this and you get that yellow. You could use Fusion to model some of the tanks, get that orange. But they don't actually have a purple icon. No Autodesk software has a purple icon, so that's going to the idea board for sure, make sure you get the whole rainbow of colors there.

      But we actually had three other tools that we could call as our other Infinity Stones. The collaboration at the beginning, the laser scanning, and the drones and the UAVs downstream for our reality captured data setting that foundation. You could argue that laser scanning and drones are one reality capture bucket that could be lumped in together, but you could also argue that the Mind Stone is a subset of the Soul Stone. But we won't get into that detail.

      The VDC Gauntlet was, using this AEC Collection, you can use all these tools. And utilized individually they've got their own super powers. Utilized together, you know, you've got the power of the VDC AEC collection with a snap of your fingers. So it makes these challenging projects almost trivial. It's not a big deal. We know how to do it. We know how [INAUDIBLE] the benefit and help out downstream.

      So we're really at the end of our time. But I was hoping if you guys have one thing you can learn today that you could take back to the office Monday and put into practice, that'd be awesome. I know everybody is different roles, so maybe you guys can start to figure out how we can ask for this type of data, how we can utilize this data differently.

      And yeah. So please, if you thought the survey was amazing, incredible, fantastic, super, let me know. If I was bad, let me know. I love the feedback. Just want to know how we can adapt for our next year. And I appreciate your guys' time. Thank you so much. Bless you.

      Downloads