Improving Efficiencies by Using Autodesk Assemble Data with Autodesk Construction Cloud Connect

KRISTIN DONALDSON: Hello, and thank you for joining us today to learn about improving efficiencies by leveraging Assemble data with ACC Connect. Please take a moment to review our Safe Harbor Statement.

Our agenda today will be introducing the team, discussing the problem, how a symbol helps, implementation at IMC, ACC Connect overview, IMC's Assemble in ACC connect workflow, how to create an ACC Connect recipe, and lessons learned.

Hi. I'm Kristin Donaldson and I am one of the product managers on the Assemble team.

SOPHAT SAM: Hi, everyone. This is Sophat Sam. I manage our integration solutions engineering team. We work with customers in terms of implementing ACC Connect across the portfolio.

DAVID MASER: And my name is Dave Maser I'm the director of construction innovation with IMC. I've got over 10 years of experience working in the industry from everything from business development, estimating, pre-construction, project management, superintendent, but always leveraging technology to provide predictable results on the projects. Certainly one of those platforms being Assemble systems for a number of years now.

So to give you a brief introduction of IMC construction, we are the second-largest construction management firm in the Philadelphia region, the mid-Atlantic market. Our headquarters is just west of Philadelphia, but we have an office in Center City, Philadelphia as well as an office in New Jersey. We've done work in Delaware. We are a rapidly growing company.

So one of the things that we take a lot of pride in is our collaborative consultative approach with our clients. You heard me reference it earlier-- providing those predictable results to our clients to give them that cost and schedule certainty. Certainly one of the ways we do that is through an innovative approach to project delivery. certainly leveraging technology to evaluate, develop, designs, construction plans to make sure that we can deliver projects in the most efficient manner as possible.

And then certainly our in-house staff is what really can set us apart, everything from safety and risk management to our in-house centers of excellence in multifamily health care, life science, et cetera. And if you go to the next slide.

So here you can see, we work in a lot of different markets. And I think that's one of our advantages is we're not a specialist company. We get a lot of different experience across a lot of different markets. Everything from a small project, maybe a quarter million, all the way to projects that are over $250 million. We do wood-frame multifamily construction, apartment complexes-- up to 250 to 400 units sometimes-- as well as major retail projects like the King of Prussia Mall expansion, one of the largest malls in America.

We've done R&D facilities for companies like Chemours that make plastics and different polymers, and certainly a number of healthcare projects, trauma facilities, MRI fit outs, et cetera. But I think what's key here is just we are a very diverse builder, not necessarily a specialist in any market.

So, one of the problems that we have is just understanding how we can collect and compile data from the models that we've been using. I mentioned earlier that we've been using VDC and Assemble and Revit for a number of years now, probably close to 10 years, but we really needed a way to compile that data.

So on the next slide, you'll see our approach is pretty uncommon in the industry. I would say what we receive all the time is 2D sketches, maybe some narratives. We will take those sketches and we will create models on our own if they're not available from a design firm.

With that we're able to get very detailed quantities very early out to the trade contractor market to make sure that we can get reliable pricing information rather than just a cost-per-square-foot number on either a curtain-wall system, a rain-screen system, whatever it may be, we're getting very detailed information at the conceptual level by taking the schematic information on the left and making these detailed, basically estimate, sheets on the right.

So, on the next slide, what you'll see is we don't just do this with any individual aspect of the project. We will do this with basically everything from building envelope, structural steel, structural concrete, interior finishes, partitions, ceilings, et cetera. Even programmatic area, whatever it is we want to be able to communicate basically the current design option, any evolution of the design, or any potential VE scenarios that the owner is considering.

We want to make sure we get these detailed quantities out to the trade contractor community, but also make sure we're communicating with the design team, with the owner, about what the expectations are and how the evolution of the design has changed over time.

The challenge with this has been, yes, these individual sheets are great as a snapshot and point in time. But to be able to compare structure option A against structure option B and tell exactly how many pieces changed, where did those pieces change, by how much did they change, looking at these sheets that can become very challenging and time consuming going line by line.

On the next slide. So, one of the other challenges that we've had is just the amount of data that we have. We have over 225 projects in Assemble. And if we want to compile all of that data to be able to compare projects against each other, we end up with millions and millions of rows of data and it takes a long time for it to refresh if you try to refresh every single row and every single column of that data.

And even if you look at an individual project, by the time we get every design iteration through a year of pre-construction we can end up with, again, millions and millions of rows of data, and it takes a long time for that data to refresh and it just becomes the same problem. You end up waiting to analyze the information rather than it being a live change and review of the design over time.

So on the next slide. The last real challenge that we had is just there's no standard naming convention for almost anything. Yes, in one project it can be easy to edit family names, edit type names. But, here again, if you're wanting to compile data across multiple projects, you have to figure out what is a simple way that you can do that. Something even as simple as wide-flange structural steel, between different design firms they name things differently. They all have their common tweaks and minor modifications, and it becomes very hard to compile those data points across multiple projects.

So we really needed a way to be able to condition the data the way that we wanted to. And on the next slide, I'm going to hand it over to Kristin. I believe she is going to talk about how Assemble can help us with that process.

KRISTIN DONALDSON: So why would you use Assemble? Assemble is a tool that helps teams ingest their messy project data. Condition that data, meaning they will want to organize, sort, and update the data so it can be visualized while making it broadly available to project teams in a meaningful way.

Well, what does that mean? Assemble unlocks and enriches the I in BIM. We enable construction professionals to condition, query, and connect BIM data to key workflows like design reviews, take offs, estimating, change management, and value engineering to reduce risk and improve efficiencies during project planning and execution. Assemble key workflows can be used across a multitude of different construction tasks.

Assemble symbol provides instant access to BIM data for preconstruction and construction workflows. We give teams access to models and views created for the project. To bridge the gap between design and construction, we give you the option to add custom properties that aid in conditioning, quantities in relevant project data with real time visualization with our 2D sheets and 3D model views.

This view has three main sections. The first section is an inventory of the project data displayed in a grid format. Then you have your 2D sheets and 3D model viewer sections that provide a visualization of the project data. The view is configurable, so you can select which sections to hide or display based on importance to your project.

We have the ability to federate AutoCAD, Revit, and Civil 3D models to allow for a more complete view of the project. By federate we mean that the architectural and structural models can be viewed in the 3D viewer at the same time to allow for a more complete view of the project.

Model-based quantities can be used for takeoff and estimating workflows. It allows you to easily group, sort, and filter your quantities; colorize inventory to manage and visualize work breakdown structures, estimates, or schedules; organize line items for a seamless push to your estimating solution; easily import your estimating coding system; verify subcontractor bids; validate against the model quantities; and allows you to easily create and save the views for your different disciplines.

Another workflow Assemble has is for automated change analysis. With each new model version, you can run variants to graphically show model additions, deletions, type changes, and quantity changes. Use variants to track changes for any numerical model property, such as area, length, or weight. You can easily identify cost implications of the changes and review quantity variances to validate the subcontractors change orders.

Assemble's mobile solutions for iPads and iPhones allows for complete site access to BIM models, data, and project quantities; offers offline capabilities for job sites look for connection; is a visualization tool for project as-builts and schedules; collects and inputs field data directly on a mobile device. And, of course, the reason we are here today, we can connect our data using ACC Connect for downstream workflows. Now I will turn it over to David to discuss why they chose to leverage Assemble data with ACC connect.

DAVID MASER: Thanks, Kristin. So the number-one reason that I would say we chose Assemble an ACC Connect to basically leverage these models is none of our estimators had to learn Revit or the details of Revit. They don't have to understand work sets and central files and the details and complications that can come with that. We're able to get the hands of our estimators on the models rather than having both our VDC and our estimators going back and forth.

We really wanted to empower our estimators to get their hands into the model, to be able to classify and connect items to CSI divisions, bid packages, the way that they want it to do that.

The next reason on the next slide is we really were on a mission, as a company, to collect and compile data across a number of our different platforms. Our client relationships management platform, our timesheet platform, our expense platform, our billing and invoicing platform, our project management platform, all of those platforms we are connecting into a data warehouse so we can really start to connect and compile data across multiple projects to really start to analyze and make some very early, informed decisions on projects that could potentially better deliver those predictable results to our clients that I referenced at the beginning of the presentation.

So on the next slide. One of the things I mentioned earlier is just the amount of data that we have. There were days where I would sit there waiting for an hour for all of the data to refresh with the traditional connector of Assemble to Power BI. And, quite frankly, that can become a frustrating process for myself or for anybody that's looking to review that amount of data over time.

So we needed a way to make it user friendly. We needed a way to, here again, get the data in front of our estimating team in a way that they could congest, it and we needed that speed to be able to do that. So now ACC Connect gives us a way to basically live-update the data as it's refreshed in Assemble. On a nightly basis, whatever the frequency is that you want to do, you can check for those updates and push it to our database to make sure we're getting the right information in front of our estimating team to make those decisions early on for the project.

On the next slide, one of the last reasons that I mentioned earlier, and Kristin talked about, is just being able to condition the data. So we've, through some time, developed basically a standard list of parameters that we've created that we require our estimating team, we require our project teams to populate. And there's a number of different ways that we're leveraging this information throughout the company, both in Assemble as well as various other data efforts, monitoring efforts, in terms of project status.

So certainly at CSI divisions that's typically populated by our estimating team. But something like the elevation ID, that might be something that gets populated either by a superintendent or a project manager, somebody that's really going to understand how are we going to schedule the project and what order of sequence do we want to schedule the different elevations. And then from there, it becomes a parameter, a piece of information that we can use across different reports and metrics throughout the company.

So on the next slide, once you have all of that information, and you have those standards, and you have that structure to the data, you can really start to create some advantages, quite honestly, in terms of efficiency. So I mentioned earlier standard and structure.

We have a number of standardized views inside of Assemble, a number of standardized Group by, like view sorts on the inventory. So here you can see grouped by CSI information, and then by the estimate item name.

Now, we've gone to looking at the way Revit wants to group information to the way our estimating team wants to group it and communicate it to the client which, like I said, you really got to sit down identify your standards, identify the structure and what it is that you want to communicate, collect, and compile over a course of a couple of projects, and then you can really start to take advantage of those in terms of structure standards and efficiencies.

And, with that, I believe I'm going to turn it over to Sophat to talk about ACC Connect.

SOPHAT SAM: Great. Thanks, Dave, and I really appreciate you providing all that background. That really helps us set the stage in terms of why companies like IMC choose to use ACC Connect for their integrations, for their workflows, and really why it's helping other customers to really ease that integration process between the different products.

So ACC connect, in a nutshell, it's really a tool that helps our customers integrate our products with other critical applications that they're using. So in this example that we've heard, IMC is using the assembled data and they want to be able to present data visualizations.

ACC Connect can do a lot more than that, certainly we'll focus on that throughout this presentation. But with ACC Connects, what our customers are able to do is literally connect our applications, like Assemble, to third-party applications. So whether that's your cloud storage, whether that's your databases, your data lakes, et cetera, the platform provides you out-of-the-box connectors to our tools as well as to those third-party tools.

Beyond that it is a automation platform that is all visual, meaning visually you're diagramming out exactly what that integration should look like. And we'll get to a demo later on just to show you what that entails. But that visual diagramming format then allows our customers to create very specific integrations between the different applications.

So even though the tools may be the same, or the integration requirements may be similar, every customer has very specific requirements, so the tool allows you to create very specific workflows that meet your business processes.

As I mentioned, it can work across a variety of tools, but the one that we'll be focusing on today is more around the data analytics by using the assembled data in Power BI. Next slide.

As I mentioned, we do provide connectors to a variety-- or to really all-- of the ACS product line. So, again, we'll be focusing on Assemble, but other customers are using ACC Connect to connect data from BIM 360, to BuildingConnected, to PlanGrid as well as Pype. Next slide.

And, ultimately, our goal is to really provide a comprehensive solution that supports all of our products within the ACS portfolio. So with the unified platform that includes Autodesk build, we also have connectors already available and immediately ready to use for our customers. Again, the goal here is to really provide a one-stop shop for our customers who need integrations between the ACS and all of that product line to third-party products. Next slide.

Yes, I'll turn it back to David to kind of walk you through how they are using ACC Connect with the Assemble data.

DAVID MASER: So one of the ways that we leverage the data throughout the life cycle of the project is communicating building elevations. On a project like this, you can see they're easily 25, 30 different building elevations that we need to communicate, not only internally to our team but externally to the client, to the trade contractors, to building inspectors, to envelope consultants. Whatever it may be, we really want to understand, when somebody talking about building elevation 2D, where is it on the building? And then ultimately, what are the materials that are on building elevation 2D?

Another thing, as you look at this building, obviously there are rotated portions, there are hidden corners, there are things that are going to be quite complicated to document on a set of 2D documents. There are going to be things that you just can't see what finish material actually wraps the corners on some of those knockouts and additional panels and corners on the facade. Is it stone that returns or is it vinyl siding? Exactly what material covers some of those undocumented hidden conditions?

But by us entering the building elevation ID into there and some of the cladding classification information that we talked about earlier, now our project teams can open up Assemble on their iPad, on their computer, start to spin the model and really understand, all right, this is the material that wraps the corner. I can take a snapshot, send it to the design team for confirmation, or we can proceed with it in the field as the model shows.

On the next slide, what you'll start to see is we're taking that information and really starting to leverage it grouping with those preset groups. Like I said, once you develop that structure and standards of the information that you're going to start to leverage across multiple projects, you can have those structured standard views to really leverage some of those efficiencies.

So now not only do you see wing A1B elevation, but you see that there's 560 square feet of fiber spin siding, 2,200 square feet of vinyl siding. Now that starts to allow not only our internal team, but also the external trade contractors, to start to understand, all right, what do I need to do in terms of planning materials? What do I need to do in terms of planning manpower, sequencing, material staging, storage, productivity?

Whatever it is, this is actionable data that people can make very informed decisions on pretty early on in the project to really start planning reliably for the next couple of weeks of the project.

So on the next slide, one of the other examples that we've started to do with this-- and this is really where we're leveraging the data from the ACC Connect tool-- is we've started to develop these standard dashboards of things that we want to look at, both internally and, again, externally. As you can see that building elevation ID is the x-axis of the graph. The y-axis being the takeoff quantity, i.e. How much of the material is there, and then the cladding classifications that we've entered in estimating to track percentage of material across the facade.

But now, not only can we start to see that every elevation obviously has different square footage, which means there should be a different schedule duration for each one of those facades, but if you look almost in the exact middle you'll see one building elevation, wing B/D, that has a material that none of the other facades do.

That is a CMU cladding classification. That means in that schedule task for that wing, that elevation, I need to add another activity and make sure that the field team, the superintendents, know like, hey, when this is coming up, I need to make sure that we have that trade contractor on site to start installing that stuff. And that's really starting to leverage some of the data to make some of those early decisions and planning efforts.

So on the next slide, what you'll start to see is, yes, I mentioned earlier that wing B had that. But now not only can I drill down into wing B as a whole and really understand, all right, maybe this elevation has 50% of the material. What are the unique materials across the facade?

Wing B was the only one that had CMU, but on the next slide, what you'll see was there was another unique material that is fiber cement siding. This material is not across every elevation, certainly. There's one that has a significant amount of it, and now we're really leveraging that data to start to understand where are the unique building elevations? What are the unique materials that we need to plan for? How much of them is there? What is the manpower or what is the material that we need to order to really start to plan and execute the project efficiently?

So, on the next slide. So, switching away from basically building elevations for a little bit, if you look at basically any cost driver of a project, you really want to make sure that you have the right scope with the trade contractors that are bidding the project. And you really want to understand that, as you sit down on a scope review, they have everything that you have.

So structural steel, as an example, yes tonnage is certainly a key driver to cost, but piece counts is also a key driver to schedule. So we wanted to be able to look at both in relationship to each other. The pink on the left is the tonnage of structural steel, basically grouped from the most to the least. If you look at the very bottom there, the facade pipe supports are something that, yes they don't weigh a lot, but they still require equipment, they still require labor to get them installed.

And that was something, specifically, that we wanted to review with every steel contractor-- what is your plan of attack to install these facade pipe supports and do you understand the true quantity that is represented by the design documents on this project? And these sheets have really allowed us to start having some intelligent conversations with those trade contractors in partnership with using the Assemble model to kind of visualize some of these scope buckets.

So, on the next slide, what you'll see is if that is one snapshot in time, for any schedule driver or for any cost and schedule driver, what we want to really be able to do is analyze how did it change over time. So if you look at the graph on the left, you'll see there's 10 or so different design options, design iterations, of this project that we worked on.

You'll see they're all tracking the percentage of material that is on the facade, the facade of the building at the bottom left there that you see, the snapshot of Assemble. the other thing that's interesting is we're tracking the square footage of glazing over time, as well as the square footage of the facade over time. So, as you can see here, at some point the facade increased square footage by almost 10,000 square feet.

Well, if you look at the snapshot of the building there on the left, you'll see towards the back of the building there is a yellow addition. This owner decided to add an emergency department to the building throughout the course of the project. The primary material on that was the EIFS, which is the pink bar that you see kind of growing in the middle, over time.

And this is really a way that we can track and communicate the evolution of the design throughout pre-construction, throughout the project. Whatever status it may be, this has been a very powerful tool for us to communicate with clients by leveraging Assemble and ACC connect.

And, next slide, please. So, both of the last two dashboards, quite honestly, if you're looking at them for the first time, they might seem very busy. And one of the big things that we're trying to communicate with our clients is, basically, change. So we decided to make another dashboard. Some people have heard me say it before-- dashboards are free once you have the data. So make something simple if you're trying to communicate a simple conversation.

So now we have a dashboard where we can pick any two different design model versions, any bid package, any CSI division. We can group it by units, whether it's piece count or tonnage for structural steel, and then we can start to analyze, what are the changes that happened between those two different design versions? And really start to have that conversation both with a trade contract, with an owner, with a design team, whatever audience it may be, this is a very simple visual to understand. These are the scope buckets that we saw change based on the models that we received from the design team.

And next slide. So the last one that we're starting to explore a little bit is being able to connect the model to our schedules to really start to be able to understand what is the productivity that we're asking some trade contractors to start delivering on some of these projects. And, ultimately, what is the variance in the productivity or the standard deviation between any structural steel sequence, any different scheduled activity. Just overall, what is the difference that were Asking

So if you look at structural steel, it's obviously piece counts per day. If you look towards the top left there, over the entire structural steel system, they're averaging about 40 pieces a day, which I think is a pretty reasonable number. But if you look at the top right, you'll see a graph that their highest sequence is 74 pieces all the way down to 17 pieces for some of the smaller sequences.

But we really want to understand, all right, why are things so different? What are the things that are driving that? So if you click, you'll start to see now with the dashboard, with the connected data, we can start to drill down on any individual sequence and understand, all right, maybe this is right. Maybe we do need a longer duration for this rather than just assuming the same duration and seeing a higher productivity.

So if you look at these two sequences of structural steel, one obviously has columns, which take longer to set. One has a canopy, which is a lot of pieces in a small, condensed area. It's going to take a while to set. And another one is the very traditional slab on deck structural steel framing for the level, which is fairly simple. You should have a higher productivity.

So between these two, yes, I can see that there would be a difference in the productivity, but they're actually backwards. The piece count is higher in the one with the canopy and the columns, so the productivity that they need to hit is higher. We should actually be increasing the duration of that rather than saying to the structural steel erector that they need to hit a higher productivity. And that's, here again, allowing us to plan better and have better execution on projects to deliver those reliable results to our clients.

And if you click one more time you'll see a third sequence. So the 57 and the 55, those two, yeah, they're very similar structural steel framing for slab on decks. There's no columns. Those two, yes, they should have roughly the same productivity.

But if you look towards the very bottom right, what we ultimately decided to do, we wanted to separate the canopy as a separate sequence, make a separate schedule activity, and really start to say, all right, that can be pulled away from the critical path. We can do that at a later date, just make sure the connections are there and available, and really start to separate it out, like I said, to make sure that we deliver that predictable results on the schedule.

And I think that is my last use case. I'm going to turn it back over to Sophat to talk about ACC Connect and how to create some of the connections with the data recipes.

SOPHAT SAM: Great. Thanks, Dave. And thanks for sharing all of that great visuals. So as David mentioned, we'll take a quick look at how we're able to get data from Assemble into a database, such as the one that David and the team are using.

So what you're seeing here is an example recipe. It's a simplified version, but hopefully it'll walk you through some of the core concepts. For example, each recipe or workflow you can create custom schedules. In this particular use case, we're scheduling this workflow, this automation, to run every 5 hours. And then I can also set specifically when it should start running. So starting at midnight every 5 hours, it's going to work its way through the list of actions that will go through next.

So the first action that we have here is connecting to Assemble and pulling view data from a particular project. As you can see, I can go in and just easily select from a list of projects that I have access to, and also select from a list of views as well. All of this as part of ACC Connect's ease of use.

What I'm going to do is switch on over to Assemble just to show you what that looks like on the Assemble side as well, just to make sure we have the data kind of mapping between the project in Assemble and what we're seeing in ACC Connects . So again, I'm going to go into that Autodesk hospital project. I'm going to bring up one of the views just to show you how the data is being mapped from Assemble into ACC Connects.

For that particular action, we're basically pulling the grid data or the column data on the left here. So this is customizable from an end-user's perspective. So you can customize all the columns, save it, use it across different projects. But what you have is all of these fields, all of these columns, available for you to then extract and pull into your own databases.

As I mentioned you can customize this however you see fit. So based off of what you want, you can just adjust that as necessary.

So going back into ACC Connect, again, this is the step where we are getting data from Assemble and from that particular view. Based off of the data from that view, we can then decide what to do with it. In this example, we have what we call instances or elements. Each one has IDs and all of the appropriate metadata or data associated with each one of those elements.

As you can see, as I'm scrolling through, each one of these properties here is based off of the column structure that we see in Assemble. so as you adjust the column structure in Assemble, you can refresh the data from that what we call output as well. So again, that's a matching column structure from Assemble to what we see in ACC connect.

Next, because we have the data that we want, what we're going to do is process each one of those elements and send it into our database. In this instance, due to the limitations of the database, we're just going to send a 100 rows at once. Depending on what you use, this may increase to up to 15,000 if you wanted to. But in this instance, we'll do a quick upsert of those elements into a database.

Here I can pick what I want to upsert into the database. In this example, it's our elements from Assemble. So I'm going to pick those batch of elements, the 100 that we selected, and now it's really just an exercise of mapping what data from Assemble should push into which field in your database. So, again, as you can see, it's all a visual exercise in terms of mapping the output from the left into the input on the right.

So depending on how many fields you want to extract, it's really just a matter of picking and choosing what you want. And, again, this is typically where we work with our customers to kind of fine-tune what's the best practices, what's really the goal of what they're trying to do, their use case, and then making sure that these recipes-- even though this is a simple one-- is scalable across all of their projects as IMC is doing across theirs.

So, again, hopefully that kind of gives you an idea of how customers like IMC is using ACC Connect to connect the data from our system into their own applications.

Next up on the next slide, I will be handing this back to David to talk to some of the lessons learned in this engagement.

DAVID MASER: Thanks, Sophat. Yeah, so some of our lessons learned. I mean, I would say the two biggest ones is really make sure you have an end goal in mind. If you can visualize the end goal, and write it down on paper, and start to identify the things that you want to track and quantify, you can figure out what standards, what data structure, you need to put in place to really collect and compile that data across multiple projects.

The other one that I would say was a big lessons learned is just make sure you take advantage of the data that's available in some of your other platforms. Things like project market, things like gross square footage, I guarantee those are tracked across every company in some sort of platform that can be connected to a database.

The more you can leverage that data and connect it between different platforms, the less duplicate information you'll have to create, maintain, et cetera, over time. And Assemble really gives us the ability to connect to our database and leverage some of that data to really make some of those early informed decisions for our clients to deliver some of that predictable results in terms of both cost and schedule certainty. And I will turn it over to Kristin to do final wrap up.

KRISTIN DONALDSON: Thank you both for that, please feel free to reach out to Assemble or ACC Connect for any questions regarding today's presentations. We thank you for your time and hope you all have a wonderful day.