Reducing Waste Through Reality Capture—Utilizing Scan Data Not Scan to BIM

MARK LABELL: So my name's Mark LaBell, I'm Senior VDC Technical Leader. I'm going to pace and run into this table a lot today because they didn't give me any room. What that role means at SSOE for me is-- it's a fancy title, it sounds really awesome-- but my day-to-day operations is helping team members, helping project people get new workflows in the works.

So we work with those project teams, integrate them, then I get the hell out of the way, let the project team do the thing. So I'm not the corporate standards bearer, just an enabler more than anything. And Tony you want to give yourself a quick introduction?

TONY PAUPERT: Tony Paupert. I am now more mixed in the projects, Project Section Manager for the VDC group. Doing day-to-day more or less overseeing the projects, being the front line to oversee and get the coordinators and specialists as we call them to adopt the same practices and do the same things so our projects have a good head start going in. So that team is key.

MARK LABELL: All right so Tony's going be helping me when we get to the live demo portion. I have about a half an hour of slides. Give some anecdotal evidence, stuff we've done on projects, stuff maybe we've not done so well, help share some lessons learned that we've done. So everybody ready for that?

AUDIENCE: Yes.

MARK LABELL: There we go. That's what I'm talking about. Got to feel the love. One the other thing, we like to run this a little non-traditionally. We do have time budgeted for questions at the end. If you've got a burning desire, a question you want answered when you see something, ask it, OK? That does help us gauge where you guys are in the room, what you want from an informational standpoint.

And we can kind of tweak it. It's not like we're going to tweak our entire presentation for you, but you know, if you got something fire it up our way, OK?

All right. So as we'd mentioned we both work at SSOE group. Tony is out of our Troy, Michigan office, I'm out of Toledo, Ohio. But I want to show you a little bit of how we work so you can understand why we work with the data as we go further into the presentation. We are a large organization, we are on seven different countries.

We do work to chase the sun on some of our projects so it's kind of a, as the sun goes one way we continuously keep working. So when we talk about point cloud data we can't just have it in a single office. It needs to be throughout our entire environment so that people can see that data fast and be able to do reliable work, or else you know what happens when people don't see point clouds. They do things the incorrect way, and things go bad really quickly.

Another thing about SSOE group is we do a lot of our work in manufacturing, so the examples you're going to see today more manufacturing-based than maybe if you're in the health care, the commercial, side. But kind of the idea still resonates with you. So think about how it would work for you. If you're in civil and you're in the land side a little bit more, you know, same type of practice we're showing you, but obviously our focus is in much of that heavy industry.

A lot of the buildings, or our clients are people that do produce the cars that you drive to and from here-- the materials that make the buildings, the food that you feed to your pets or yourself, cereal-- that's what we do. You know, computers, that's our clients is the people that do that type of work.

Also the standard of care that we do is we exhibit that we want the highest quality deliverables to our clients. So as part of our standard of care is doing not just engineering and architecture work and startup, but we also do procurement and construction management as well. So not every job we do run that way. So once again, these are some of the ways that you're going to see that's kind of the reason why we've pushed it this way.

Upfront apologies, as I said I'm an excitable guy. I'm very passionate about this, all right. I do tend to rub people the wrong way on the Scan to BIM stuff. Don't get up right away, hear me out please, OK? I'm pleaing with everybody here that I know, as an industry we can do much better with how we treat point cloud data, reality capture data, whatever you may want to call it-- that flavor of the week word-- we can do much better with it.

So don't everybody go start rioting because you're not happy with how the polls went. You know, got to get a little humor in here. So I'm apologizing, you know. Point clouds are my passion. I've spent a lot of time over the past couple of years working on them, lot of investment. So don't mean to offend you, but if you don't like it just let's stick around and hopefully I'll convert you in the end or at least get you to think a different way.

So we are a very lean organization, or we try to be lean on several of our projects. And anybody want to try and define lean? Not talking about your goals after the 1st of January after everybody eats on Thanksgiving. Anybody want to give a crack? One happy person in the back.

AUDIENCE: [INAUDIBLE]

MARK LABELL: There you go. It's really simple. I mean, lean.org, it's pretty much that. You know, it means creating value, aligning your goals with your client, not just trying to be a profit center and make tons of money. But a lot of times as AEs, we keep on trying to hammer the same big box of shit into a square hole. Sorry, I'm very passionate as I said.

So talk with your clients, understand your client's goals, adapt your workflow and processes to them. That is truly lean.

So a little bit of story time. This helps understand why we are the way we are, what we're going to present later on. So our first project, when we actually started doing Laser Scan, started way back in 2008-- business leaders over there on the left, I'm on the right. My artwork is really awesome, by the way. I'm good with point clouds, I have third grade-- I think my five-year-old daughter draws better than me now.

So he comes up to me, he says, hey, we got a project we want to scan-- technology guy. Awesome, let's do it, sounds great. Didn't have a plan. We'll talk about that in a moment. So I get excited.

Yay, we're going to do this. Awesome, we really didn't plan it very well. They scanned it. Data's too large, 2008. Anybody work with point clouds in 2008? Anybody not agree with that statement? We had to buy a supercomputer just to be able to view the data about 10 to 12 grand investment. But when you have 10, 12 people on the project you can't buy 10, 12 computers with no forethought and no plan. So you can imagine where that went.

A ticked off a business leader, after five weeks, hey, we had to go scan to BIM, because in 2008 that really was the gold standard. I mean that was the only way to really look at point cloud data is you had your real views or your picture views maybe, and then you had to just convert the minimalist model stuff to work with the data.

So we did that. It put a little bit of a hurt on our schedule. End of the day, client was still happy. They were able to immerse ourselves into the facility that they weren't intimately at. That's a big thing, but we could have done better.

We did lose a little bit of money, but happy client, good relationship. Overall not bad.

Early 2010, you think that two years later we might have had a better plan in process, but project manager there on the left, I'm on the right. The project manager walks up to my desk-- I'll never forget this snowy day in January-- 30% due by the end of the week. Here's a hard drive full of data.

Talk about a plan there. Anybody ever have a hard drive just show up magically at their desk? No? One, two. OK. Yeah, that was awesome, by the way. You can imagine that was my response.

Did we get something together? Yes. It took seven people, seven highly technical staff members three days to cobble up the biggest pile of crap that still made them happy, but seven people plus the design team that still had to get the documentation out. So very bad planning.

Early 2013, things are getting better, you know, we're starting to involve it a little bit more. I'm on the left, C-suite's on the right. We now understand that where we are in the industry things have caught up, and they become a little bit more mainstream. Scanning is becoming more democratized-- I know that's an Autodesk word there-- but, you know, it's becoming more accessible, it's becoming cheaper, hardware is more readily available, it's starting to actually go directly into our suite of programs.

So we have a small project, very tight spacing, trying to get pipes from one end of a plant to another. And it was a great job, because it was just basically chase the pipe rack down and make sure that you know-- scan it. One day of scanning. Very small cost. $2,500 just to scan that job. It is an easy job.

C-suite still said no. $2,500. Now think about this, we're going to send engineers in the field with tape measures and put them in a scissor lift. It's not very safe behavior, and probably going to take longer than $2,500 hours of fieldwork. But this is tried-and-true method, remember. Everybody still knows how to grab a tape measure and rely on that data.

It's worked for 30, 40, 50, 60 years. Why do I need to trust the tech group? The first two projects that we had in eight and 10 weren't so great. We remember that. Now we've had some success on other jobs where clients drove the process, but they still didn't believe we can actually do it as a non-driven thing from our client.

Ended up we paid for it out of our department, went well. We learned a lot of things from that. So then the general manager of construction-- I don't think he actually joined me today. I'm looking around the room, he said he's going to be here.

He's a great guy to have on your side. He doesn't actually lay back ever in his desk like that, but it's just fun to put that there right, you know. And he calls me-- I remember, this was right before I had my youngest daughter-- says you know we want a scan of this project. We owned it. You know, this is a [? CM ?] at risk job. This is our job to screw up. We don't want to screw it up obviously, but we want to scan it.

Now we've obviously had some successes, we have people on bigger jobs that know how to work with the data, but we don't have quite enough people at SSOE now that are trained to work on the additional influx that we're going to take on. So I actually fired back and said for once, we're going to actually hold on this one because we don't have enough people. Nothing worse than going to spend 20, 25 grand on scanning a job you don't use the data.

That's bad. You know, client's going to throw egg at you, you're not going to get a good relationship. So we ended up just doing it traditionally, which turned out to be kind of a bad thing because 70-year-old facility, how many of those are really built to the as-builts if you can even find them. Anybody want to take a guess?

Those columns were two feet off which is bad. You would never think structures that far off but in the 100 plus jobs we've scanned in the past two years I can probably tell you on 25% of the jobs, we're getting two to four-foot busts on columns on these facilities. It's crazy. You would never think that.

Two-foot bust, they started pre-fabbing all the steel for the structure, for the equipment. We ended up losing $200,000 out of our profit because of that, because of that one simple mistake. Now everybody knows the saying about the flavor of bad quality and such. Well this wasn't bad quality, we just lost on this one.

But from a construction management side, we still procured and bought out packages and still got everything installed turned up and compressed the schedule by several weeks. Anytime you can turn a product line on several weeks earlier means they're making product faster, which means they're making money. They're happy still.

So like the previous story in 2008, we still got a happy client, we just lost money again. We don't like losing money. I don't think anybody in the room likes losing money I'm pretty sure.

So what happened there was quite possibly the best thing we could have ever done for scanning at that point. The GM realized he wanted to do it. We weren't quite trained well enough. We lost money because of that. He changed the mind of the C-suite, and this is what happened in 2014-- and this is no joke.

So 2014, the water, the hose turned up. And it's not just me, there are plenty of people that do this. Tony has worked with this from a Revit standpoint. We have plant people that are invested in this, plenty of other scanning people that work with this. And by scanning at SSOE, we don't own a scanner.

This is a completely vendor network. I can call up vendors whenever I need them, they have as many scanners as I need to get jobs done as fast as possible. We've scanned 1.5 million square feet in less than four weeks. 2,200 scan positions, I believe, using [? survey traverse ?]

So present day, where we're at today, I put the Oprah thing in here because it's true. It's just like her show. You get a scan, get a scan, get a scan. Two weeks before I came here, no lie, seven proposals we wrote up for scanning. None of them less than $20,000 scanning PO cost. I mean that's a relatively big job in terms of scanning, and when we scan we only get the data.

We don't care about anything else. You scan it, give me the raw data, and we'll walk through that process here in a moment.

So anecdotally, and this is where I come off a bit brash. Just listen to the evidence of the most important man, because in '14 we also converted 700,000 square feet of a manufacturing facility in Scan to BIM. It took months to do, and it helped us with coordination mildly. But if we would've just learned to work with a point cloud data, once again, would have saved money.

And we were using automation tools that you'll see in a little bit. So that's why we say this.

So project example. This is a particular project where we gained a relationship with a client. You know, you start doing small work, that's how client relationships go. You get the trust, you start getting bigger work.

So project number one, you know, we go in and they're going to tell us what to do because we don't have quite the relationship yet have to say, we're going to change your game for you. So, short project, very low cost, issue for constructions, very tight deadline. I mean everybody deals with this. These are things that we know happen on our job.

The thing that was really a pain in the rear from a goal standpoint is we had to stick to their CAD standards which-- CAD standards. Does CAD standards really-- can anybody argue this-- Does CAD standards really make a project more successful in terms of construction?

OK. I'm glad the room agrees with me on that one. So what we knew when we went into the project, obviously, like we said, schedule driven. OK, good. DOT work. So they have a deliverable that the government needs to have at the end of the day that matches that look and feel.

That these facilities, all of it's scanned TIFF files. You know, Mylar, onion skin. Some of it's been converted to CAD, some of it's mismatched. They're using Raster Design to overlay CAD on hand drawings. It's a frickin' mess, OK? And the drawings aren't updated like anything else we've already dealt with.

You go in the field, you know, there's 10 pipes that are either missing, that aren't in the drawings or vice versa. There's 10 more pipes that just shouldn't be there.

So the client direction. We want it to look like a TIFF, once again going back into the smashing the turd into the box. Some of the 3D design was done in MicroStation. We do MicroStation in-house as well. We're not just an Autodesk firm. Whatever our clients tell us to work in, we got to work in it. I mean, at the end of the day, like everybody in this room, if you're going to get paid for something you're usually going to try to figure it out if they're going to tell you to do that.

And bills of material. So this is the piping example, more heavy process piping. All the build materials hand calculated and counted from the 2D drawings, because once again we're looking at TIFFs and all that stuff.

And then to make the job even crazier we were the engineer on that job. There is your typical GC role, and then the client purchased everything less than two inches. So they still needed to detail it and understand it, but then they purchased it. So it really knew the contracting very hard.

So uncovered. Kind of already said this, but we had TIFFs. Of course they're rotated, because when they scan them in anybody that remembers scanning TIFFs they start twisting in the scanner. The scales get all jacked up. So even if you bring it into Raster CAD things get really kind of squirrely quickly.

Line continuations from piping didn't match from one sheet to the other. Even though it said go to sheet two for the continuation, sheet two didn't show the pipe. Awesome. It really makes it a good deliverable set when you're trying to take it off for estimating or actually building the place.

No typical details or consistent material specs, which no consistent material specs and anybody in the plant and piping world makes it kind of hard. Plus the design level of detail was completely outdated so a lot of people in the office were much like the "Y you no like me" meme, and very frustrated because it was very hard to adapt to what we normally do as an industry-- a very cut and dry design, deliver, very detailed documents, easy information and then hand it off. It was very, complex.

So this is one of the example drawings that we dealt with it. It looks pretty good, huh. Anybody be able to build off of that? Any contractors in the room by the way? You really think you want to tie 10% bid to that? Do you feel comfortable with that, especially when you can't read the right hand side of that screen there?

So you can see the clouded work over there on the left. That was our new scope of work. That was our design, by the way. You had read that too. Now luckily there was a sheet that had a blow up detail so the guys that raised their hand again, you feel comfortable 10% getting close? Of course you're going to pad your bid, and that's the standard of how they run all their jobs.

I mean that's just how they work. I mean, most companies still run like this, most clients. And this is why we didn't inform them right away. We wanted to work with them. Now here was a detail. We still detailed it in 3D and then we had to smash it down into 2D and get it back in Raster CAD. So still high level of detail. This was a lot better for you, but not an easy process to put all the dots together.

So project one made money. Everything was OK, but way too painful. So never forget the phone call on this one as well. Myself and a couple other technology people were called into a room with the project engineer, and he said how can we change the deliverable strategy. We've been given projects two and three. They're not even going to bid it out. We did good enough on project one, which that's the goal of every job, right?

You don't want to be bid against your competition on the next job because you want to be able to get handed work. We've achieved that now, and they gave us jobs two and three right away instead of phasing them out. So it was like, all right, we get more work and faster. how do we get better, because we can't do the same thing again.

So here's an example of once again what you're looking at from project one. You have your CAD there that was detailed out, and then you have your Raster overlay that we had to give out for bidding. How do we match-- remember the project goal, client deliverable. They want that look and feel still, and that's what they are giving for bids.

How do we match that, but we don't need to give that to bids any more. We're going to give them a real design set, an actual set of drawings that looks like you would normally do. All new stuff, some existing modeling, bills of materials that are accurate, not hand-counted. And something that you can actually build with confidence, not looking at these crazy rasterized drawings.

So our pitch was, we're going to give you your as-built, you know, DOT-regulated files, but after construction's done. We're not even worry ever about doing that. That was a kind of a radical mindset. Took a lot of conversations with that client to get there.

So the new process was scan every site. And by the way, each of these projects has about five or six different job sites that you have to do, so when we say two projects there's 10 sites, 12 sites. So a lot a lot of work. Highly agile team, constantly doing work processing.

We model it in 3D, we have our ISOs generated from the 3D model, bills of materials all automated. Obviously, it makes sense. Most of the room I don't have to argue this why we do this anymore. Eight years ago at Autodesk University we would have been talking about how awesome this is, but now we're in a better place.

And then when it came to construction documents bids we were still given a 2D drawing package like you normally would, and that TIFF is as part of that as-built process once again. So as you can see here in the upper right there's a scan of the facility. Here's our new equipment skid with the piping going directly to that.

Notice that we didn't do Scan to BIM on there. There's the point cloud and we designed right to it. So for a refresher here's the original drawing that we delivered. This was our proof of concept to get the client to buy in. There's the same 3D model. Now we didn't detail all the dimensions, because we told them we'll dimension it, but are you OK with that? Looks the exact same.

They were happy with that. So kind of summarize how those two projects went. Project one is on the top right now. We'll bring up project two.

So scope definition within the organization, you know, getting our team aligned. That's the first thing you always do on a project, hopefully. We do a drawing request, so we would actually request all the drawings. And that process, as you can see, spans past 30% review.

We don't even have their drawings at 30% review on project one. I forgot to mention that actually. That made it even trickier. So at 30% review we're going in with napkin sketches and PDF redlines. Imagine what your project's going to look like.

That was the process. So then you get your drawings received right before 60%, then you try to jam all of your napkin sketches into 2D drawings that look like garbage already. Nothing really reflected it so you're still going to 60% with napkin sketches, still a lot of uncertainty.

Contractors on board at that point. They're already starting to prep what their estimation is going to be off of 60% design from napkin sketches. Not easy. Can be done, but very time intensive. Then you go through your 90% and it was bid phase off of those documents.

Rewriting the playbook on the bottom. We would scan it, we would get the data. So they you're scanning the sites, it'd take about four days. We would walk off-site on Thursday, we would have the scan date in-house on Monday and working it on Tuesday in Navisworks and Plant 3D.

It's pretty remarkable turn around, because they scanned it. We still haven't even requested drawings yet, and we're already doing our design modeling in 3D. So huge game-changer. We're ahead of the curve actually on the job.

We're actually doing 30% design reviews with the client fully immersed in their site, which by the way, like I said, all these sites, they're spread all throughout the country. It's not like they're all within driving distance of where they're at.

Just like any of us, very immobilized workforce. They never get to see these sites. That was a huge advantage, because now they don't have to fly to all the sites and do walk-downs all the time because they were doing walk-downs at every review.

So we'd still keep the walk-down at 60% where we'd go there but we would also do a model review with that. And then same thing through 90%, but we were also giving these drawings-- we were taking screen shots, and we'll show that here in a moment-- to the contractors for very critical pieces of information. long-lead items for equipment, long-lead items for special pieces of pipe that take longer to get in.

So a much different set of documentation is part of that process. How are we looking on time? OK, good. So here's the initial scan of that particular site. You can see all the pieces of equipment that we went through and demoed. Of course the middle swath, all that equipment's gone.

What is hard to see based on the projector here-- it looks a lot better on a high-res monitors. On the left-hand side the piping's missing. The top of the piping's missing, and you can see here on the right if we tumbled it, the piping's missing there.

So we've cut that all out of our model, no longer there. Why do we do that? Well, one, you don't have to worry about doing in Clash Detection on piping that's not there when you know you're going to demo it. Two, it makes it easier for the person that's actually going to do it from the modeling perspective, and know what space they have.

So once again as we had stated before we do a little bit of Scan to BIM. The green pipes there are showing the existing connections. You do need that in your drawings for ISOs and other reference material. So we do spend a little bit of time tracing those things, but once again, we're not doing the whole thing. Very specified.

Also when we hand these models over-- because with this client we also added the models over to them. Everything, full transparency now. And they're extremely excited about this. We let them know that anything that's in an existing model-- this is important, so any of you that try this method make sure you document this-- existing models don't use that for dimensioning when you're going into construction.

The point cloud is the gospel. That's the most accurate thing, OK? Don't use the existing, because we've had one or two that came back and said, well you modeled it, that mean it's the thing we're supposed to dimension off of. No, use the point cloud. The column is in the right spot.

You can't completely model columns 100% accurate. They have twists, they have flange bends, forklifts, any of that stuff hits them.

So here's our new design. Once again, that piece of equipment, just a little bit different view of what we're looking at. And here's some of our deliverables now in 2D format. So much more detailed than what you saw before. A pretty similar piece of equipment, but you have your ISOs, your one-lines, your build materials. All of that, once again, is completely automated. We don't talk about too much of the praise of that.

Here's some of the construction reports as we did as part of that. So not only do we give a traditional bid package out to the contractors on this job, we went in gave them an additional step. Here's those critical pieces of equipment. We threw this in.

We're using a plug-in called iConstruct for Navisworks to do these views. Put in these comments and then we send that out as part of the additional bid package. So they have better information of what's there. You know, what's the critical piece. You know we know that from an engineering standpoint, but they don't know that yet because they may not have been in the game completely.

Just another view with another kind of overlook of that. And just a different part of that. Once again, you can see there the green pipes re-tied into the existing, all the white pipes and the valves, all brand new work.

And this is how we modeled them. We were cutting our 2D views from them. What's the question?

AUDIENCE: [INAUDIBLE]

MARK LABELL: This particular job, no. The point clouds don't show in 2D The workflow that we're going to pull up later, yeah. We are putting point clouds in 2D. So that's actually the workflow. Were going to show that specifically with Revit, but we can do it with Plant as well. It requires a little bit of trickery. It's not easy.

So from a case-study standpoint and the successes on this job, we were able to do the same, or actually more work with less people. And it's not like we said, OK, we got two, three less people to do the jobs so we're going to fire the. No, we put them on the other job that we want as a part of this.

So we were able to take it on and say, no, we can't do that type of thing. Provided much clearer construction documents. The bids were coming in tighter than that client has ever experienced before, because the package was much more complete and not just looking at old hand-drawn sketches basically.

And of course at the end we still met their deliverable, their client requirement, the look and feel that they still needed at when it went to as-built.

Another case study. This is more of the materials that you're going to see, the particular project that we're going to break up here in the live demo in a moment. This job we ended up winning it because we came in with an aggressive schedule. We're going to basically reduce it by five months. That's big when you can do that on a particular job.

They had it drawn out as kind of weird, but we showed how we can help provide value in that process. We, from a demolition standpoint, using that demo process, 14 RFIs and 400,000 square feet of retrofit.

14. Three of which related to structure that wasn't built to any AISC manual, so they had, of course-- you know, that's going to happen.

New construction's still going on so I can't give you an RFI in totality there to say, is it trending lower, better. But we know it's there, but just is it, you know, what orders of magnitude.

But from a VDC BIM person in the room here. I'm not sure how many there are. Actually let's do a show of hands. How many [? have been? ?] 15%.

All you guys do Clash Detection? Do you guys love Clash Detection, especially all the ones that are false? Nobody's shaking their head on that one, right? By doing demolition point clouds. Now it might seem crazy that you're going to spend time in trace where you're going to demo stuff out, 50%.

So when you're getting in 400,000 square feet, this just the particular area we'll show it a little bit, you'll get maybe about 800. We were getting almost 2000 clashes per area, and each area was, let's say about 40,000-60,000 square feet. So that's a lot of just BS that you're cutting out of your daily job. So that's a big deal there.

So it might seem like you're expending a lot of effort to do this at first, but there are a lot of other benefits. Obviously clear documentation, highly accurate deliverables, but also the benefit to you as what your job role may be or somebody else in your organization.

What in the devil did I just do? OK, there we go.

One last case study. So this is one that we actually did a full Scan to BIM on. So like I said I'm not completely against it, but once again, do it where it makes sense. This particular job, 50, 60-year-old piping, completely corroding, sagging, twisting, bad shape. Cut it off at the walls, and basically replace it. That was the scope of the job.

We want the same exact thing that's already there, don't need to re-engineer it, but you know, we want to also be able to prefab it and bring it in. So they're going to put 80 hours of field work to measure everything, document it, get all the valves, take all that information. They spent six hours total using a hand scanner. Three of it was in a car, and they scanned it twice just for good measure.

74 hours saved just by scanning it. We use that time to train that person in Plant 3D, they had not used it before. Brand new engineer to the organization. So we use that as a benefit saying, OK, this job was originally going to be done in 2D much like the other job. Hand takeoffs, hand build materials. We weren't even going to do ISOs because it wasn't required.

They just wanted a set of drawings to bid, the contractor was going to hand it all off. So on this job we did a fully detailed 3D model. We had all the ISOs build materials. We still saved them 100 hours on a time and material budget, and we saved them additional money on the contractor not doing all the detailing as well, which of course awards us more work. And you know, it's a good thing.

So here on the right-hand side the reason the cloud looks so weird with the colors is because we've overlaid the 3D model directly on top of the hand scanner data. And of course anybody that's used a hand scanner, it's photometric, or light sensitive. It's not quite as good as what you're looking at in LIDAR. So it does have a little bit of a weird ghosting effect. Yes, question.

AUDIENCE: [INAUDIBLE]

MARK LABELL: Right, thank you.

AUDIENCE: [INAUDIBLE]

MARK LABELL: It depends on what we model, and we don't determine that until we walk through the process.

AUDIENCE: [INAUDIBLE]

MARK LABELL: Yeah, the guy that learned this for the first time and you know, mind you, we had Qubit. We don't even use Qubit anymore because of the integration of-- this was done in Plant 3D, 2016. Some of the new nodal snap features, he used it in there. I mean, he was learning how to use the program, but within I think five days, he had it pretty close to buttoned up in terms of all the XYZ location. But that's with the learning curve too.

I mean he's a he's a pretty intelligent kid, he caught on really fast so.

AUDIENCE: [INAUDIBLE]

MARK LABELL: Yeah. Yeah. So any other questions? OK. So what changed, why did we take the approach that we did in the industry? Solid state drives. Huge, improvement to what we deal. Without those bad boys we would not be able to process the data that we do.

And one of the things that you'll see here in a moment is we replicate all of our data across all of our offices in the US meaning from a point cloud standpoint, we have about four terabytes, roughly, of active project data that replicates between all of our offices. That's a hell of a problem, by the way. It's not easy to do. We won't talk about that because I can spend two, three hours just talking about what not to do.

Now what we do is localize the data so yes, it's replicated to every office, bit now it's replicated to the user's workstation. But replication is not the right word, but we'll use it for simplicity here. So if the data is on the network and that person works on project x they have the ability to pull it down, and that's huge performance impact.

We're talking one terabyte drives in over 200 people's machines at SSOE, and we'll have a return on investment in a month. We have the proven data behind that. One month for one terabyte drive.

Also what happened is recap, big game changer for us and as well as everybody. We're able to get the data directly into our Autodesk software. We weren't using something crazy like a-- I can't remember the Cyclone product or something else. That made it a lot easier to integrate directly within there, so those two things.

And if you look at the time frames of when both of those came out that's right around 2013 when we started pushing things because we realized at that point the software was there and the hardware was there to be able to complement how we work as an organization.

So what do we use from an application standpoint? Now across the top is, let's call it our typical workflow-- 90% to 95% of our jobs we're using Navisworks, we're using Plant 3D, we use Revit, it we use Inventor, we use Civil 3D, pretty much everything, OK, at that point.

We are a large organization, and we have a lot of different business units that do different design features. So not every job uses all those soft-wares, but I can tell you right now we have a lot of jobs that do use Revit, Plant 3-D, Navisworks, and ReCap. That's very common at SSOE and it's not easy to do. And of course there some of the others.

Lessons learned. Of course you can see 2014 across the bottom in yellow and then the teal and the blue. 2015, 2016. We spent a lot of time getting people to be able to train it in the modeling software. Biggest bang for the buck obviously, right away. What we found, profoundly, is if you don't engage the engineers to look at the point cloud data and train them on how to use ReCap software, guess what they're going to do? Anybody want to guess?

They're going to go back in the field and measure, which is going to be lost time, lost money. You know, a huge waste especially, I mean, a lot of the field work we do, not really within driving distance of an office.

AUDIENCE: [INAUDIBLE]

MARK LABELL: Oh, yeah.

AUDIENCE: [INAUDIBLE]

MARK LABELL: Yeah, so what happened in 2015, as you can see on the right there, point cloud viewing. That's your engineers, that's your project managers, stakeholders, estimators, basically not your modeling staff. What we would do when a project kicked off hour-and-a-half required training if they had scans involved in that job.

I don't care if you're never going to look at it because you're an instrumentation guy you're sitting through the training, because you probably will use it at some point on the job. And we do go through those tools of Pipe Snap. And yes, some of the tools early on with the ReCap team, that confidence wasn't there. Now there's a very contextual-- and obviously with JetStream it's gotten better as well.

You know, the confidence is there, but the first couple of jobs, you. You got a guy who's been there for 30 years or a female that's been there for 30 years, I don't trust it. Rightfully so. It's hard to change everybody's mind. Yep.

AUDIENCE: [INAUDIBLE]

MARK LABELL: So we have a vendor network, they're required to do that. So most of them-- not to sound biased towards a product-- but all of our vendors use Leica equipment. Kind of hard to say not biased, so they're required to do other registration. They do it all on Cyclone.

So we had to, of course, as we've gotten more proficient we've developed custom apps. On the left-hand side there you see our application that actually copies down and monitors all the projects. So it's not just like your copying the data. If you change project from project x to Project y you can copy the next one, and if somebody adds new data to it, it alerts you right away. Full-alert system.

And then on the right-hand side, yes, that's ReCap, but all of those tags are column markers. So you can just plug in a column grid, say where you're starting at. Column A-1, you're going all the way to Z-36, tell it the offsets, and bang, all of your columns are offset.

That particular job had 1,200 columns. The application was written in a few short hours and it saved obviously a lot of money. So that's something we use pretty often.

One other thing, based on time I'm going to cut it, is Navis [? to points ?] because we will talk about it. But this is a crazy workflow that we engineered to actually reverse engineer 2D and 3D data in Navisworks to get it back into ReCap so we can actually do our demo process.

And everybody looks so freaking confused that I just said that, so I think we got a good cliffhanger at the moment. And one last thing we'll talk about is, there's a new hack in ReCap. So we unify all of our scan data and we put it into logical sets, and we'll go through that in a moment. But you also have your ReCap project that has all the Real Views associated to it.

So you have two sets of point cloud data, which makes your network storage even bigger. And for us trying to pipe it across the world makes it harder. So what you do here is you delete all of the points. You can still see the bevels are there. You do a "Save as." OK. Remember this. "Save as."