Revit for modular design, prefabrication and repetitive layouts.

BRIDGET WHITE: This is Revit for modular design, prefabrication, and repetitive layouts. Just to make sure you're in the right spot. I'm Bridget White. I'm Architect and BIM Manager for Nordic Office of Architecture. We are 144 employees from 27 different countries, and we specialize in airports, schools, hospitals, and government facilities.

KRISTOFFER TUNGLAND: Hello everyone. My name is Kristoffer Tungland. I'm a MEP Engineer and a BIM Manager. I work for COWI. We are a 6,600 employees in 81 offices in 24 countries. And we also work with airports, hospital, government facilities, and bridges, tunnels, and marine structure, and many more.

BRIDGET WHITE: OK, the class objectives for today. We want to give you guys insight into a large scale Norwegian hospital with the proposed industrialized construction; show you how you can evolve a traditionally built Revit model into modulation; learn new modulation and duplication workflows in Revit; streamline the modular information flow; and simulate construction sequencing.

So the project Stavanger University Hospital, also known as SUS 2023. Stavanger University Hospital is situated in Stavanger, which is very well known for its oil production. The project came about due to the centralization of hospitals in Norway, the pressing need for research in hospital facilities in the region. It's 100,000 square meters in its first phase, to be built on-site in 2023, and the second phase is 100,000 square meters in addition.

So the project makeup, we have four main buildings, which are centralized around a central courtyard. Each building has its own courtyard as well to increase day lighting for patient recovery. There is a central axis for transportation. And the four buildings are connected underground and on levels two and three by a transparent glass bridge, or ring as we call it. And this is to make a transparent thoroughfare between buildings.

Our project team, we're really proud to be here representing our project team today. We're a big team, and we are led by Helse Stavanger. We're two architect firms, two engineering firms, and a landscape architecture company.

OK, SUS, the project structure and deliveries. We have a very ambitious project ahead of us. All disciplines, obviously, working in them and complete digital information transfer. We're aiming for a complete lifelike visualization, accurate quantification from the early phases, and paperless delivery, essentially.

So we're using augmented reality-- BIM Cave, Oculus Rift. Wind, energy, daylight, environmental, and performance analysis. A lot of this is actually consultant-based, a lot we're doing within the project team. We've got database synchronization with dRofus. This means the client can inform the project through an external database. Quantification through IFC, and 5D construction sequencing. Yeah, and paperless, as I mentioned earlier.

So here we have the BIM Cave. It's a 1-to-1 representation of our project. And this helps the doctors and nurses understand the thoroughfare from building to building. It's a very large project and sometimes we get sort of a lot of comments. My office is too small. It's too far from here to here. And this gives them a little bit more insight into the project.

The project file structure, everyone's in Revit apart from one. One company is using particular. We're all linking between one another, and we have external consultants inserting information through fire and acoustic demands.

What's important to note here is that the architects divide per building, and we separated out the facade file and the rank file. The other consultants are working within one project file for the moment, as long as the file size holds out. But we also have the demand from the Norwegian government that stipulates that we need to transfer files in IFC. So this has kind of formed some of our workflow, which you'll see a bit later on.

OK. The conceptual phase. And when I say conceptual phase, I'm not meaning methoding. We're well into the project where we're making layouts. We're checking the functionality of the program. But we're still in that conceptual phase with things are moving, things are changing. At this point, the project was actually under a traditional build. So we weren't considering industrialization at this point of the project.

OK. But if we take a slice of the second floor, I can say very early on the repetition in the project is very high. And this can be brought down to a few factors. Being a hospital project. Of course, the wards essentially are duplicated many, many times. But the interesting thing to note here is that the repetition varies per discipline. The architects have a repetition in the bedroom wards. But the MEP, where it's structural engineers, have a different type of repetition. Maybe in the zones or the duplication of electrical features.

And this we handle per discipline. What suits us per discipline we manage. But seeing as I'm an architect, I'm going to show you the architecture aspect of it. Here in three of the buildings, you can see, bedroom wards. And the repetition is actually within the bedrooms. So if you take one of the buildings and kind of analyze that repetition, we can say that the red sections are one type of bedroom ward. And the orange are the mirror counterpart. They're exactly the same, just mirrored. And we have some small deviations-- the bedrooms for spatial types of wards.

So in our early phases, we handled this with groups. And the nice thing about groups, which is kind of important for our later progress, is they have a base point in the group. Which means that they understand their relativity within the project. They understand where they're sitting. And we can flick fleck other groups in and out of the project. And become a little bit more efficient with making new design layouts.

We can also create group libraries. Revit holds a memory of all the groups that you create. And when we need to change the plan layout, we can actually go back and forward and flick fleck in and out, ward layouts. And make new concepts quite quickly. One thing I want to note here is that, when you're kind of creating a strategy for your design duplication, it's important to assess what your delivery demands are going to be in the future. This can affect the way you group. It can affect the way you organize your groups. Groups are really nice. But once you start getting into a grouping and design options, you can come across some really nasty warnings. Which means that you have to escape from your grouping methodology.

One example I can give you is grouping grid line to grid line. This was a nice way of grouping for us because we had a seat grid. And we could flick fleck in and out ward layouts and very easily. But, because we're delivering quantification in those early phases, we ended up with half walls. And we had to make calculative changes to make sure that we had correct quantities all the way through. And obviously seeing as we separate the facade file, we're using room separation lines to contain the rooms. And this can also lead to a few hundred warnings, if you're not careful with what you're doing.

I think this is nice actually. We would have benefited from a place holder in those early phases. We know that we have 200 of the same. We're not too sure what they are. But we need to kind get the ball rolling and get the functionality of the plan working. But we don't really need as much detail as we could have. We could have come back on the detail label. And that's something that was worth us kind of pointing out. Early evidence of repetition in the facades. Obviously, being such a large project, we have 50,000 square meters of facades in the first phase.

We were using a traditional curtain wall workflow, with quantity take offs from the curtain walls-- which was quite nice. Some global parameters. Some intelligent curtain panel walls. And using design options to fleck through kind of a normal process we use in that phase for facades. Separating the facade fall helped a lot, controlling the likeness of the facades and the five buildings. But one drawback can be is, if you're using a facade analysis programs, that some of them don't like if you separate the facade out in a separate file like that. So just be aware of that issue.

OK. Moving on to the developer design phase, where the concept and industrialization became a major factor in this project. I mentioned earlier that Savannah is a region of a lot of oil production in this region. But in the past two years there's been a downturn in oil production, meaning a lot of highly skilled engineers and industry workers have been left unemployed. And this is a really nice collaboration actually. The region needs a kind of a new configuration of work. And we have a lot of highly skilled people sitting around. And so we're collaborating with the oil industry in this project to create an industrialization strategy.

Why else would you do industrialization in SUS? Well, cost factors. It's cost-beneficial. We can get a better product for a less price. Our quality assurance goes up. Building efficiency goes up. Controlled conditions for building off site. Speed of production goes up. And it's opportunity for innovation. We're reducing waste. And we're increasing accuracy. And we're also creating a wide diversity and skill level. Using an industrialization strategy, we can actually employ a wider range or skill basis. And have unskilled workers in the kind of assembly or installation phases.

OK. So then, we had to find opportunities or methods for industrialize acing the project. And what we did was we created a potentials map where we could create a list or oversight of potential areas where we had a mess of duplication. And there was a potential to create kind of intelligent objects or groups of parts. And from those potentials map, we used, obviously, a methodology to analyze the potentials. And look at the constraints and opportunities of the potential modules or potential strategy.

And some of the constraints for this project were module size. In Norway, we have a very smaller module size that we're allowed to transport. We have medical gas, so we cannot modulate medical gas due to regulations, meaning we need it all checked site and assembled on site. And we had a pre-approved layout, which meant that we couldn't place additional structure and kind of ruin our nice little plan that we'd made in that phase before. At this point, after we had come up with a strategy that we were happy with, we actually took contact with Bryden Wood-- who are professionals in prefabrication industry in London-- to get kind of an insight into how exactly this could work on site.

How we could cope with like a very steep production phase. And how we could build these four buildings and even out that kind of a work on site. They've got some very exciting projects, so I recommend you check out their stuff too. And alongside this, obviously, there's a change in information flow. We can no longer produce the same information that we produce in a traditional project. There's new types of information requirements. And this required a new BIM process development plan. And we call it process development because it's ongoing. It's evolving. It's changing.

But we needed to align with our information requirements in this type of project or strategy. OK. So module identification. This was kind of like the potentials map. Here we take a piece of the plan. And we assess it for duplication and the potential to industrialize. Here we differentiated the facade from interior modules, just because of kind of a workflow difference. And we begin to select the modules that we're considering for fabrication or industrialisation. After that, we could isolate our modules.

And here we have our modulation strategy, which is actually just change last week. I didn't tell Kristoffer. But we changed a couple of modules last week. So it's an evolving process. But yes, we can identify the modules within that kind of isolated area of plan. And we can start listing or making kind of a list of parts or modules which we're intending to fabricate. And in this project we have a horizontal module, which is an attachment under the slab and the hallway, which is an intelligent ceiling panel which incorporates MEP.

We have a vertical module, which is structure and MEP. We have bathroom modules. Intelligent door modules. We have furniture modules, which also take interior linings. We have facade modules. And intelligent plumbing modules, if I can say plumbing module. OK. So the BIM process development project. This is when we looked at our modulation strategy and decided that we couldn't work the way we wanted to work. And we needed to come up with something new. So that we could form this project and produce the right level of information.

Our main goal was obviously the seamless cross discipline control of the modulation strategy. Take better advantage of duplication. Present the strategy through IFC, being in Norway. Align the project outputs to the building industry demands. Produce the right level of detail at the right time. And have the ability to simulate modular construction sequencing. How is this going to be assembled? What kind of docking? We need control over that, before we kind of let this strategy out for tender.

But we also need to take care to allow ongoing design and visualization workflows. Allow for project analysis. And maintain a workflow suitable to all users. My team at the start included 13-year Revit users and they got a short, sharp shock into using Revit. But we can't then introduce something that's really complex. And they can't manage their day to day workflow. We have to take it into consideration, the level of our users. So we began organizing the project. We formed a modular team, which is different to the usual per building team we have.

And the modular team was made up of a mock per module responsible. Because obviously we have the same modules over all four built-for buildings. And there was a dedicated BIM responsible for each module. Then we formed the typelist. And this is a cross-discipline way of naming and coding our modules. And how we can cut code and label our list of parts. And yes, I say there was cross discipline. So our information is translatable between disciplines. We understand which part is part of which discipline. And we can start breaking down each module and coding that as well. OK. Now's the fun part.

KRISTOFFER TUNGLAND: Thank you, Bridget. I'm excited to tell you how we made this work in Revit. We found out that we had to have two workflows, separated. We have the modules. And we have the facades. So I'm going to start with the module workflow now. See on the left. The module workflow is a way to handle duplication of industrialized and cross discipline modules in Revit. And handling parameter values of all elements within the modules.

So the purpose is to avoid repetitive modeling on similar modules. And fill out information only once. To do that, we introduced a model with placeholders for all the modules. I'm going to get back to that and we created a tool for placing modules. And fill out the Palmetto values. And the tool can also audit a model and check it for changes and update them. I'm going to go back to that, also. So here's a short animation to show you the concept that we thought of. We want to insert placeholders in a shared file.

We share that with all of the participants. And then we want each discipline to use this file to place their part of the module. So each discipline has a group in their file. And defines that in their file. And places it on the same spot as the placeholder. So if the placeholder moves, we want it to be audited and moved after, if you want. So to do this, we had to define three new Revit models. We have three Revit models to do this. And we start with the placeholder model. And that model defines the module location and the type coding.

And the discipline model-- this is where you normally work. Do your data work. And that is the model that contains the part of your module as a group. At last, we have the model that defines the complete module. I'm going to go back to all this. So this is the placeholder. It represents a unique module type. And it has a geometric outline of the module. So we can share how big is this for the team. It holds type code information. You place that on this place holder. And it's developed early in the stage.

So it's easier. You can find out how many modules to have in the whole building early and can do a cost on it and other thing. It's a separate mass family. We decided to use a mass family in Revit. And it's developed by few team members. This placeholder communicates the module types to the team. So the next one is the placeholder model. It defines the location of all the modules in the project. And it holds instant information of a model, like it's unique ID. It's also developed early in the stage and by a few team members.

And it communicates the module instances to the team. So the next one is the normal working file, The display model. It links in the placeholder model. And has been part of the module as a Revit group. And the groups are named according to the type list, as we showed earlier. And only dedicated, disciplined members work with this groups. At last, we have the module model. It contains individual modules as link from all disciplines. And it's formed by each discipline saving out their module groups as a separate file. And it's linked together in a complete module.

This file can be used as a complete model for prefabrication. You can send this to the guys who want to make that one. And to make all this work, we have developed a tool for inserting an element, group, or assembly at the host element location. So in this case, the host element is the placeholder. The tool has a filter to find desired elements. And elements can be in linked files or in your local project file-- the current file. Inserted elements, they get a reference to its host element. So we get the idea of the host element into the inserted. So you can track it later on. And it can copy parameter values from host element to inserted element.

So we can get the type coding from the [INAUDIBLE] to the group and all the things inside that group. Because we store the reference to the host element, we can audit model for changes. And update both location and insert values on the inside of the element. So the tool started as dynamo scripts. But we needed a better user interface to use it. And it was a hard process. So we decided to make an add-in to Revit instead. So now I'm going to show you a little demo of the tool in action, if it works.

So now you can see the model with all the placeholders. And I'm going to run the tool. Here you have configurations of all the modules on the left side. And you can select which file you want, which categories you want. We use the mass in this example. And we can filter out the elements we want. So this is the name of the mass object that we want to find and insert. Insert the group. You can see you can select what you want to insert on the line there. And then when we're doing a refresh, you will get a list of all the placeholders in the model.

And when we press the single, it's going to create an instance of that group on every single spot. So back to the model. I can see it inserted it inserted elements. You can also use this on repetitive design. It's the same technique. I'm just going to show everyone the thing is inside there. Created 15. And if something changes in the placeholder model-- now, I'm going to move a place holder. And this one. Normally, you will have the placeholder model as a link. This is just easy to show you how it works. It's a bit jumpy. Bad connection.

So now I'm going into the tool again, . I'm just doing a new refresh. I'll find out that one instance needs to be updated. And just press update. Also, we'll do that on parametric values. And it's moot. So we have some challenges. The MEP duct and pipe connectors inside the group. It's really hard to do. You can't connect them, so that's one challenge. To do that, we thought of having a connection node. So we will place a node or a connection object in the group. And we can use the tool you saw to find all those nodes. And installed a new node in the file. So you can draw from that, long and out.

So I think that's more like how they do it in the real world. They have like a connection. Like for electrical, it's a connection and you have to define it. And I think that's a good workflow to do it. And then we take all the calculation values on the module in the group. And just copy over the values to the connection node in the model. So you will get the calculated values on the connection in the model. And also, you need unique values on some parameters. So we have to be sure to set-- values can vary by group instance, if you're not in Revit. And to be sure you can have a unique idea on the toilet there. Or else it will be controlled by the group itself.

And also we have some limitations. We can get the rotation of the group through the API. So we had to make a workaround. So maybe the module gets rotated in the wrong direction. We have to rotate it back and accept that change. So we hope we can get the better support for that. And also, we have another one that's extensible storage. We save all the configurations on the group. And if you change the offset or change their reference level, it will lose the extensible storage group. So we have to keep that and reapply it. So please help us.

[AUDIO OUT]

BRIDGET WHITE: Yeah. That's a good question. And actually, in this project it's a little easier to [INAUDIBLE] intelligent parts. And at this point we have [INAUDIBLE]. But we're looking at [INAUDIBLE] as well. [INAUDIBLE] But in the conceptual phase, obviously, [INAUDIBLE] But in terms of like connections and stuff, maybe--

KRISTOFFER TUNGLAND: I also want to point out that the placeholder-- you defined outline. So you have some guidelines to when you're modeling. And you can do it fairly early in the project. So all the other members that are modeling around your module are going to know where that exactly spot will be.

AUDIENCE: And did you also-- a follow-up question-- did you also recreate-- we had a lot of problems with hosted objects that were part of groups that were not hosted in the object that was not part of that group. For example, I saw that you had a whole bunch of plotting objects or whatever. Some of those could be wall hosted, but the wall might not necessarily be within that same group, therefore creating a problem.

KRISTOFFER TUNGLAND: We never used the wall hosted elements, just free.

AUDIENCE: [INAUDIBLE]

BRIDGET WHITE: Yes. So we didn't have that problem [INAUDIBLE], yeah.

KRISTOFFER TUNGLAND: Face and wall hosted elements, we tried to not use them.

AUDIENCE: So how many projects does it take for you to get to this point? Obviously, I don't think that you started this process of the groups and all that on this project. Maybe you did. But how far down the road were you before this project with this process?

BRIDGET WHITE: This is the first project [INAUDIBLE]. But we had used it before [INAUDIBLE] design. So we know how this works. We know the limitations. And we now have a tool for it [INAUDIBLE]. But in terms of [INAUDIBLE], it's the first time we have [INAUDIBLE].

KRISTOFFER TUNGLAND: And the tool was used before other project, but not done for this purpose because we used it for like ventilation, need connection for electrical. That was the idea to create the tool. So we found out we could use that in this one also.

AUDIENCE: My question is with regards to how far have you thought out the concept? Everything [INAUDIBLE] seems to be geared toward the design of the project. Have you given much thought about how you would want to integrate the contractors into this type of workflow? And/or [INAUDIBLE] this type of workflow has on your subcontractors as they try to digest the model?

BRIDGET WHITE: Yeah, well, it's paperless, so they will be using our IFC models. We're trying to move them into more of a digital. But essentially, they will perhaps create their own models in the end as well.

KRISTOFFER TUNGLAND: And then we just swapped--

BRIDGET WHITE: Swapped.

AUDIENCE: Is that your long-term game plan to kind of keep the design separate from the construction site? Are you expecting your contractors to generate their separate model or were--

BRIDGET WHITE: Well--

AUDIENCE: --were you hoping to integrate them into the project?

KRISTOFFER TUNGLAND: I think both.

AUDIENCE: Both?

BRIDGET WHITE: We would like not to. We would like it to go straight through to-- straight through, but it's not possible at this point. Yes?

AUDIENCE: [INAUDIBLE] those contractors into the design, the set design. But I just want to elaborate on it. It's currently challenging in the same, coming from a contractor. We receive the design, but it's only backgrounds, and we would rebuild using proprietary families. And then we would create groups or assemblies for premanufacturing.

BRIDGET WHITE: Yes.

AUDIENCE: And extract that metadata to ensure that that duplication or replication [INAUDIBLE] doesn't get out of control to bring them on ahead of time. So that way, the design crosses over the manufacturing seamlessly without having to say, hey, Contractor A wants to do it their way, so they're going to change your model, which could create unforeseen conflicts. Therefore, [INAUDIBLE] resolution.

KRISTOFFER TUNGLAND: They need to follow some sort of outline. And the connection points needs to be on the same spots. So you have define something for them.

AUDIENCE: OK.

KRISTOFFER TUNGLAND: In the back?

AUDIENCE: [INAUDIBLE] between 50% and 60% of the project that comes with an architectural project because they're not built for modular [INAUDIBLE]. They did mention [INAUDIBLE]. With that correction, I would like to ask you, what do you think [INAUDIBLE] help out those 50% to 60% of the companies who work [INAUDIBLE] being rejected for not being [INAUDIBLE] modular. What advice would you give them? What should we think of first and foremost?

BRIDGET WHITE: You're meaning-- just so that I clarify the question-- you're meaning they're rejected because--

AUDIENCE: They can't be built.

BRIDGET WHITE: Right, right.

AUDIENCE: [INAUDIBLE]--

BRIDGET WHITE: Yeah.

AUDIENCE: --there's so much demand for them for a number of reasons you identified, but they can't be built. [INAUDIBLE] design.

BRIDGET WHITE: We're lucky actually in this project because of the collaboration with the oil industry. And we have a lot of-- we have a lot of workers in Stavanger and a lot of competence in Stavanger for this particular methodology of building. So we have energy for it. And it's helped this project move very quickly towards this kind of strategy.

Advice for other projects? I'm not sure actually. But look to other companies that are innovating with modules. And the stuff that we looked at with Bryden Wood is extremely inspiring. And there's so much potential actually. It's just being the first to build it is a little scary I think.

AUDIENCE: I have two quick ones. Just for clarity, who created the [INAUDIBLE]? Was that the [INAUDIBLE]?

KRISTOFFER TUNGLAND: No. I started with the diamond script.

AUDIENCE: OK.

KRISTOFFER TUNGLAND: And we have two developments in [INAUDIBLE]. So we had done the hard stuff on it. And we have collaborate with others also. And we made it ourselves.

AUDIENCE: And then you mentioned pricing was part of that. What did you use for pricing [INAUDIBLE]? Was that Bryden Wood or somebody else?

BRIDGET WHITE: No. We're currently doing that now. And it's a very difficult process. But it's actually down to a bunch of parameters because if we can present the modulation strategy and we know how many modules of how many types, then they can kind of start to form an idea of the price. But they don't know. We have to work very, very closely with the industry on that pricing. But it will be a ballpark, an educated ballpark figure for that.

Yes?

AUDIENCE: One question. The toolbar that had the coordination tool on it, I noticed it had a lot of tools on it which I had not seen before. Are those all plug-ins that you guys created?

KRISTOFFER TUNGLAND: Yeah, yeah. It's for the company.

AUDIENCE: So [INAUDIBLE].

KRISTOFFER TUNGLAND: But the team is using the tool.

AUDIENCE: [iNAUDIBLE]?

KRISTOFFER TUNGLAND: But we will hope it will be included in Revit soon, some sort of, yeah.

BRIDGET WHITE: Yes, go ahead, yeah.

AUDIENCE: How do you account for a change in structure about each module [INAUDIBLE]?

BRIDGET WHITE: A change in strategy, for example, like the modules change.

AUDIENCE: Structure--

BRIDGET WHITE: Structure--

AUDIENCE: [INAUDIBLE]?

BRIDGET WHITE: Yeah.

AUDIENCE: [INAUDIBLE]?

BRIDGET WHITE: Yeah, how do you manage a change in the structural elements?

KRISTOFFER TUNGLAND: Do you think if something is bigger than higher some places? Yeah, we have to do one module for each type. So we haven't done it parametric. But the coding has the idea to have it almost the same code, just another letter for this height and another for this height.

BRIDGET WHITE: But it is an issue. We have different floor heights.

Yes?

AUDIENCE: Could you do groups within the model if you link them, if you link them from outside?

BRIDGET WHITE: You're talking about the discipline model? The discipline model is essentially the placement of the groups. But the placeholder tool, or the placeholder file, stipulates where those groups will go.

AUDIENCE: And the groups themselves, are they groups within Revit model, or are you linking them--

BRIDGET WHITE: No.

AUDIENCE: --in separate modules?

BRIDGET WHITE: No. They are within the Revit model. And we need them for other analysis factors within the project mean that we need to model within the discipline model.

AUDIENCE: So what evaluation-- because we track both, and we saw benefits and drawbacks--

BRIDGET WHITE: Yes.

AUDIENCE: --of both systems.

BRIDGET WHITE: We need to deliver to IFC, and we cannot export the links and the places. If they have different--

AUDIENCE: [INAUDIBLE]?

BRIDGET WHITE: Yeah, exactly. So when you export and it's a link, it'll lose its connection to your coordination system and export to a local zero on top of one another kind of thing. But what we do do in the module model is link. Link the other disciplines' groups together. So we do link there. Yes.

Yes. And thank you for coming.

[APPLAUSE]