Just a decade ago, delivering a model instead of drawings was a new idea. As the practice became more common, it has changed the very shape of the built environment. The current state of model-based delivery is a snapshot of a new technology being adopted, one that could help us understand how best to integrate future technologies. For our 2014 Annual Report, we brought together a structural engineer, a steel fabricator and two construction managers to discuss how model-based delivery has changed – and will continue to change – the way the industry works. Here’s what they had to say.
From left: Serge Dussault, Frank Falciani, Steve Hofmeister and Allan Paull. (Danny Bright photos.)
Serge Dussault, B.Sc.A., M.Eng., P.E.
Vice President, Engineering, Canam Group
Serge has more than 20 years of experience in the design, fabrication and erection of steel structures. He manages a number of Canam’s design and detailing offices.
Frank W. Falciani, MBA, LEED AP, CCM
Senior Vice President and General Manager, dck worldwide
Frank has 40 years of experience in every facet of the construction business. He manages major construction projects with a focus on providing clients with the best return on their investment.
W. Steven Hofmeister, P.E., S.E., LEED AP
Managing Principal, Thornton Tomasetti
Steve has 30 years of experience in structural engineering for complex projects delivered via design-build and design-assist methods. A member of Thornton Tomasetti’s board of directors, he oversees the firm’s structural and construction engineering practices.
Allan M. Paull, P.E.
Senior Vice President, Tishman Construction, an AECOM Company
Allan has more than 34 years of experience in the construction industry. He heads review and oversight of structural design and construction for all Tishman projects in the United States.
Delivering the model has become increasingly common. What are the biggest benefits?
HOFMEISTER We started sharing and delivering models about 11 years ago because of the limitations of 2D drawings. All the intelligence that we generate about the structure that we generate over eight or nine months of design is difficult to convey in a set of 2D documents. Delivering a model provides a better level of information, a more complete view of our knowledge. A model can also be better understood by other members of the project team.
PAULL The 3D environment more clearly shows issues that have to be resolved. Take a simple thing like the length of steel members. If we are running off 2D drawings without connection designs, the subcontractors don’t know the exact length, so they go from working point to working point. But we know that the braces could be ordered three to four feet shorter. Just the material savings of having accurate lengths that account for connections is a monetary savings for the clients.
DUSSAULT As fabricators, what we like about model delivery is that at bid time we all have the same understanding of the project. The project is more fully defined. I think it levels the field in terms of what to expect and what we have to price, which in turn becomes a benefit to the owner. There is much less interpretation of documents and the scope of work that has to be priced is much clearer.
HOFMEISTER When we share models or have models as a deliverable, we see contract prices in a much narrower range at bid and negotiations.
DUSSAULT Correct. It eliminates bidders making a mistake, which unfortunately happens – not on purpose, but it happens. Often the one making the mistake at bid time is chosen to do the project. The model reduces the risk of error, so the owner gets several prices that are much more representative of the project.
HOFMEISTER It eliminates the low-price bids. There’s no ability to miss the complexity of the scope. It’s there. It’s very visible. The owner’s bid-day price might be a little higher on some jobs, but it will be closer to the right price.
DUSSAULT Yes. The second big advantage is in schedules. With 2D drawings, owners and general contractors need more time to analyze the bids and the pricing, so it takes longer for them to release us on a project. And once they release us, we’re usually already late, right? A well-built model gets us off the block much faster.
PAULL The biggest benefits for construction are predictability of price and schedule. People are realizing that the design-bid-build model for complex structural steel isn’t producing the best results. If drawings were perfect, we could send them straight into the shop for fabrication. But in reality, a lot of work is required to get them into the shop. So for highly complex jobs, it’s important to start shop drawings earlier in the game. When we do, we get much more predictable schedules. We avoid extras, delay recovery plans and things of that sort.
HOFMEISTER Yes. With a 3D model, we’re now able to overlap design and construction activities that had to be sequential before. We can do much of the early construction-phase work in the model during design. The design-assist approach allows us to work directly with the fabricator, erector and general contractor and get their input during design. That knowledge is then included in the model we hand over.
A Revit model (left, click to enlarge) of the complex platform being built over active rail yards to support Manhattan’s new Hudson Yards project (right). Most of the development’s 12.6 million square feet will rest atop the platform. Thornton Tomasetti performed structural engineering as well as steel modeling and connection design for the platform. Delivering a detailed steel model reduced change orders, simplified fabrication and shortened the schedule for the platform’s complex structural steel. We are also providing engineering services for several of the new buildings. (Model ©Thornton Tomasetti; Rendering courtesy Courtesy Related-Oxford)
FALCIANI From a general contractor’s perspective, we look at the structural and architectural models as the baseline from which we’re going to develop our pricing and schedule. It helps us clearly see what it’s going to take to build a project.
Another benefit is being able to predict the full impact of changes. Schedules are all-important, especially for stadiums and major shopping centers. The deadlines can’t change. But owners want the flexibility to make design changes throughout the process. Changes either take time or cost money. A complete model – one that includes 3D, 4D (time) and 5D (cost) information – allows us to test proposed changes. We can price them and see where the impacts are. And we can clearly show all of this to owners, so they can make informed decisions.
HOFMEISTER Designers and fabricators are also able to react more quickly when changes occur. A model enables us to incorporate, analyze and deliver the changes much sooner than we could before.
FALCIANI I agree completely. Modeling does a lot more than just clash projections. It can help with many aspects of a job. Having that clear information available – to owners, designers, subcontractors, everyone – is incredibly useful making decisions.
HOFMEISTER I think we accomplished the clash detection in 2004 or 2005. That was the low-hanging fruit. All these other benefits, that’s where we are now, and where we are headed is improving the process even more.
Are there any drawbacks?
HOFMEISTER Some people will say there are a lot, that there’s no reason we should ever do this. Other people – like me – believe that we’ve always had the responsibility and the liability potential. The real risk on the design side is that if we hand off a higher level of information, any errors are “not right at a higher level.” It puts a greater onus on us to deliver a more complete product. If we don’t, there is more risk.
Steve Hofmeister and Serge Dussault. (Danny Bright photo.)
DUSSAULT I agree. It puts a lot of pressure on the professionals who deliver the model. If it’s not right, it’s going to come back and cause a schedule impact, a cost impact. So designers have to do more work in a time span that hasn’t increased proportionately.
The quality of the work we receive can be an issue, because this process is somewhat new. It has been happening for about 11 years, as Steve said earlier, but in the construction industry, that’s what we call “new” – we’re not in the software business! So model delivery varies a lot from firm to firm. Right now, there’s no standard. We’re not always sure what we will get.
HOFMEISTER Professional organizations such as AISC and AIA have been working to standardize modeling techniques and technology, but they can’t really standardize what our deliverables are, or what you need from us. In the years we’ve been doing this, it’s all been evolving. Frankly – and I might be biased – I think Thornton Tomasetti is ahead of many others. But it’s not a level playing field yet. Look at drawings: a set of our drawings will look very much like a set from one of our competitors. Over the past 100 years, they have become very standardized. The models are not there yet.
DUSSAULT Not there at all.
HOFMEISTER Over the next 10 years, we will start to see the models morphing towards a common form. The content, quality and level of precision for models will become much more similar, like drawings have.
PAULL Well, I’ll say there are no drawbacks, but it’s a dual model. We find that it has to be actively managed on our end, just like another contractor. We have had success using various levels of models, ranging from a geometric model all the way to full detailing. We did a job in New York for the Port Authority, the World Trade Center Vehicle Security Center, with Thornton Tomasetti, and we fully detailed the job. We never gave up the detail and it went very well with lots of changes, and we were able to move to a very complex structure. But there has to be someone actively there, managing the information.
There are no real negatives, because you’re starting the process earlier by doing it in the design process, which gives you a bigger bang for the buck. But it doesn’t happen by itself. In detailing, little things can grow into big things because of the information flow. Simple things, like where a hole goes. If that’s missing, you don’t know where the next hole goes, and then you don’t know where the next opening is. Before you know it, you’re totally mucked up. So it’s very important to have somebody from the CM contractor’s side actively managing the model or interfacing with the structural engineer who’s doing the work. Like anything else in life, it doesn’t just happen on its own.
HOFMEISTER Some people would certainly say there are negatives and there are risks. But I think they’re ignoring the fact that the process has always been imperfect: the architect has 20 consultants, the contractor has 20 subs, and there is an owner or multiple owners. It’s complicated. The models are just one way to improve that process. I think you’re exactly right: there is no real negative.
PAULL Clients often worry that having the designers make a model is a waste of money. They think the steel contractors are going to do it all over again because they have their own way of doing things. But the reality is that complex connections have philosophy and they will go the way the structural engineer wants them to go. And for simple connections, everybody does about the same thing. The trick is to start earlier, while the drawings are being developed, so you can find problems early in the game. The cost of any errors and omissions then is much less than after steel has been ordered and time has been scheduled. Or, God forbid, it hits the field.
HOFMEISTER That’s the worst.
FALCIANI I agree completely. There is no downside, but there are challenges. And one is that the subcontractor market is not sophisticated enough to keep up with the modeling that is being presented by the architects and engineers in real time. I’ve written over 800 bid packages and I’ve tried different ways to get subcontractors to demonstrate that they have the expertise to participate in the modeling process and provide real-time information in to the CM throughout construction. Allan is right: the CM has to manage that process, and it’s only going to be as good as the information coming in from the subcontractors.
We need to be able to update the model to make real-time decisions about schedule, cost impacts and what changes are going to be made. The subcontractors fully acknowledge that they need to keep up with the model and update it. But getting that to happen in the field has been a challenge, and I don’t see it improving as quickly as we would like. To compensate, we end up doing a lot of that work for the subcontractors because they can’t get the information in at the speed that we need it. The more advanced major subcontractors have that capability, but of course, that spawns an entirely different set of issues that must be addressed.
PAULL We have a job now on which they are going to do a model concurrently with the structural design before we buy the job, and the owner is keeping the modeler on his side of the ledger. He is not going to let the modeler become the detailer for the steel contract so that the model can be managed. It essentially becomes an advocate for defending changes.
HOFMEISTER You both just brought up a really good point. The value modeling brings and how smoothly the process works has a lot to do with the team. If the subcontractors on a project aren’t of the right caliber, there is a business risk for us. But when we’re working with subcontractors that are team players, the modeling really helps the design process sing. That’s when it can solve problems early and clarify the “what-ifs.”
DUSSAULT Yes. It depends on the players. Not everybody has the same experience with model exchanges. It also depends on how the contract is set up, how the model is purchased. If this very sophisticated tool is being purchased as just a product, with limits on the roles of the designer, subcontractor and fabricator, there will be problems. If the model is not delivered properly, or if there are flaws, there can be arguments about responsibility. But if the model is purchased more like a team member, it can bring the team together. The model can create better teamwork on a project, with big benefits for the schedule and in a reduced number of disputes and RFIs. We’re also able to get to work much faster when we have a good model that is being handled by a good team.
HOFMEISTER On average jobs, about 60 percent of shop drawings come in at least twice. A few years ago, some of us here at the table worked together on MetLife Stadium. On that project, we cut the numbers of RFIs dramatically, and 97 percent of the shop drawings were accurate on the first pass. That’s a huge benefit to everybody involved.
PAULL Yes. We started doing geometric models seven or eight years ago, and they reduced geometric problems on the job tremendously. It also showed us that the next critical issue was the connections. We used to go out with a set of 2D drawings and it would be an ordeal because of issues like members not sized, members not located properly, missing dimensions or dimensions not closing. So there’s a huge advantage in getting all of that information up front, and when you have it in a 3D format, you can see where everything is going. It’s very hard to visualize members moving in three dimensions in a 2D format.
HOFMEISTER 3D modeling definitely gives you the ability to see more.
How will the shift to model-based delivery change the roles, relationships and responsibilities among owners, architects, engineers, fabricators and contractors?
PAULL My opinion is that detailing should really be separated from the structural steel contractors. They still have a lot to do: fabrication drawings to do; erection drawings; scrubbing the drawings. But I think the information needs to be on the design side. This can lead to another problem, which is that erectors and fabricators and detailers look for certain things that designers don’t. But I think putting the detailing on the design side of the ledger, or the owner’s side, or buying and managing it as a separate consultant produces a better job. That lets the steel contractors do what they really want to do: to fabricate, to hit the shop windows, to erect. There is still a lot of specialized engineering for them to do. But I think it’s kind of backwards for them to be searching for information or guessing what something should be, or putting in allowances or making assumptions for an estimate or a bid.
DUSSAULT Allow me to disagree a little, Allan. If it’s a simple job, if there aren’t any sophisticated materials and we can buy anywhere, it’s OK to get shop drawings and just fabricate. But on more complex jobs, the shop drawing process allows us to plan the job and to buy the material right. If we skip that part, it’s going to be more difficult. We’ll have to adjust, if that’s where the market is going, and so we have been participating a lot in model exchanges.
HOFMEISTER We found early on that the modeling technology isn’t the real challenge. It has improved a lot in the 11 years since we started doing this together. We’ve also had to change the mix of our design team, to add people from the construction side who understand more of what the fabricators and erectors need us to deliver.
DUSSAULT We also prefer to keep control of the shop drawings themselves, so we know when they are coming and when they will be ready for fabrication. If we’re not an active a participant in that process, then the planning of our work becomes more challenging. If there are items with a long lead time, but we don’t know about them until receive shop drawings, it can cause delays.
HOFMEISTER I think you both raise good points.
PAULL I’m obviously kind of opinionated, having been through a lot of cycles.
HOFMEISTER We all are. That’s why we’re sitting here, having this conversation. This reinforces my opinion of my favorite procurement scenario, which is when we team with the subcontractors early so they are involved in the entire design process. It does two things: we get that critical input from the subcontractor; and the subcontractor develops the necessary knowledge of the project.
DUSSAULT True.
HOFMEISTER That’s the best of both worlds, when you have the team together, much as we did on past projects such as Yankee Stadium and MetLife Stadium.
Tekla model of Yankee Stadium design. (©Thornton Tomasetti)
PAULL It costs a lot of money to model a complex building, so clients can have a little sticker shock when they see that number. There’s some resistance from dollars, but it’s the right thing to do. It produces a better outcome for the clients. Fees for the designers have to go up to provide that fuller service, but I think it is money well spent.
DUSSAULT As long as the owner doesn’t expect a simple equation of spending an extra X dollars on the design consultant and saving the same X dollars on the subcontractor. That equation doesn’t work out to a net zero when you just look at the bid. You need to look further downstream, at the coordination that is done up front, the schedule that’s saved, the reduction in RFIs: that’s where the real savings are in delivering a good model.
FALCIANI Absolutely. I work with a lot of high-end shopping center owners and developers. They tend to look at the expense of the model the way they view sustainable construction: “How much will it cost, and when is the payback? What’s the return on investment?” If you can convince owners to look at sustainability over the long term rather than just the initial costs, there’s usually a large return on that investment. We have to convince clients that the same thing is true with modeling a major construction project. There is so much to gain. The model is a management tool throughout the construction process. It can also be used as a defensive tool to prevent cost overruns and to help challenge any claims that may arise. The information available in an effectively managed model is a huge benefit to most clients.
The model is a very good tool whenever claims are submitted. A model that has been updated and maintained shows the progression of construction against the schedule. It can be used to quickly develop time-impact analyses to show the subcontractors, so they can rethink their position before we all start paying the lawyers. I have always been an advocate for the client. My job is to save as much of my client’s money as possible, so we always assume a defensive position first and focus on the project. Having an integrated model with the schedule and cost built in makes that job better and easier.
PAULL Another interesting thing: Clients seem now to be more focused on schedule. They want to know when the building can get built, when they can start operating. As they learn that a model can speed up the detailing, fabrication and construction, that seems to be driving the decision-making – more so than even the money of the model.
FALCIANI They all want faster delivery. There’s no question about it.
PAULL And the biggest risk on a job is the steel.
Will the adoption of model-based drive changes in overall project delivery or increase the use of emerging delivery methods like Integrated Project Delivery (IPD)?
The AIA defines IPD as “a project delivery method that integrates people, systems, business structures and practices into a process that collaboratively harnesses the talents and insights of all participants to reduce waste and optimize efficiency through all phases of design, fabrication and construction.”
The IPD delivery method uses multi-party contracts to incentivize team members to collaborate toward an outcome that is successful for all. The AIA also recognizes an IPD philosophy that can improve any project, regardless of contract structure. Advanced modeling technology and model sharing are critical components of the collaboration necessary for the full spectrum of IPD projects.
While industry organizations describe IPD with differing levels of detail, fundamental concepts are shared:
Mutual Respect and Trust: Team members rely on collaboration and teamwork to support the best interests of the project.
Mutual Benefit and Reward: Compensation rewards behavior that’s best for the project. Early involvement is recognized and rewarded.
Organization and Leadership: Team members commit to common goals and values. Leadership is assigned to the team member most capable of specific work or services.
Collaborative Innovation: A free exchange of ideas among team members promotes creative decision-making. Ideas are judged on their merits, not the role or status of their originator.
Early Involvement: Early involvement improves decision-making. Diverse knowledge has greater value if it is employed earlier.
Early Goal Definition: Project goals are developed early and are agreed upon and respected by all team members.
Intensified Planning: Increased effort in planning drives efficiency and savings by streamlining and shortening the construction effort.
Open Communication: Direct and honest communication focuses energy on quick identification and resolution of problems, rather than liability.
Technology: specify technologies up front to maximize functionality, generality and interoperability. Open and interoperable data exchanges are essential.
DUSSAULT It will cause a greater level of collaboration. I think IPD is a little like what BIM was seven or eight years ago: everybody has their own definition. The details of IPD still need to become clearer, but earlier and better collaboration is what gets a better product – better not only in terms of quality, but also in terms of price.
HOFMEISTER If you read the IPD documentation, you’ll find a few basic behaviors the team must commit to (see sidebar). I’ve never done a “true IPD” project, but some of the very best projects of my 30-year career have had those behaviors: the teamwork, the collaboration, the open discussion of challenges and issues. You don’t need an IPD contract to have those behaviors, but if you have those behaviors the project will go better. We’ve all witnessed that.
FALCIANI I’ve been on projects where we’ve taken pieces of the model – particularly in hospitals, where we have those congested mechanical systems that go above the ceilings – and used them to have these materials prefabricated in a shop; literally bring them over, put them up in the air with a roustabout, quick-connect them and move on. So there’s a huge benefit to integrating project delivery if you can get sophisticated subcontractors that are able to work with you hand in hand.
HOFMEISTER To answer the question, I don’t think model-based delivery will drive IPD or any other delivery method. It won’t drive, but it will definitely enable. And it might even persuade some people to adopt more of those fundamental collaborative behaviors.
PAULL We’re also seeing more architects working in a 3D environment, and I think it’s naturally heading in that direction. In construction, no one wants to be the first to experiment, but someone has to move the process forward.
HOFMEISTER In most of what we do, success comes down to how people work together. No one entity – in this huge process of 20 or 30 consultants and 20 or 30 subs – has all the information. We have to collaborate with the rest of the team. Modeling gives us a better tool to collaborate with.
PAULL As you’re seeing more “cloud-based” models, it’s going to become easier and easier.
HOFMEISTER That technology has really improved.
FALCIANI The industry’s definitely heading that way.
PAULL Some things are just too complicated to do in 2D. A great example is One World Trade Center. We had a high density of rebar at the bottom of the building. No one could figure out how to get all the steel and all the rebar in. So we sat in a room with the rebar detailer, concrete contractor, steel detailer, steel contractor and structural engineers. In just two or three hours, we were able to move the rebar around in the model and get everything in. That’s something that would have been almost impossible to do with paper drawings.
HOFMEISTER You used the model as a collaborative tool.
It’s clear that model-based delivery can reduce overall costs for owners. How does it affect the bottom line for engineers, fabricators, construction managers and contractors?
HOFMEISTER As designers, what we really sell is our knowledge and our ability to convey that knowledge. If we deliver more information per project, it takes us more time and effort to produce them, so we need to get paid for that. Clients haven’t always understood why we need the additional fee – and they haven’t recognized that they were paying it anyway. The most obvious way they pay is in the bid, when the steel fabricator’s detailing puts that knowledge in there. But there is another cost that comes after that, in schedule delays, when you’re going back and forth with RFIs…
DUSSAULT RFIs, yes.
HOFMEISTER …and going back and forth with all the discussions around job standards that drive schedule up and cause the subs on the construction site to ask for more money.
Over the 11 years I’ve been involved in this, I’ve seen owners and contractors develop a much better understanding of where that cost really is. Then they’re willing to shift some of that money around. I also know that if I hand Serge a model that takes things up to a certain level, he can’t just take what I give him and move forward. He needs to back up a level and verify it to fully understand the job. So his initial cost doesn’t go down as much as my cost goes up, but at the end of the day, the owner’s total cost will be lower than before.
Thornton Tomasetti has performed a number of “true IPD” projects, including an ongoing hospital expansion at the NorthBay Medical Center in Fairfield, California. (©Thornton Tomasetti)
DUSSAULT You cut so much dispute – and cost – when the information is right, when you reduce RFIs and there are fewer questions throughout the project. But is it more profitable? It all depends on the team you’re working with. Model delivery is a tool we’re using to speed up the process and engage in better teamwork. So if the team is the right one, yes, it’s going to be a more profitable project, and not just for us, but for everybody involved. And the team is not just the structural fabricator and the consulting engineer: it starts with the owner and goes all the way down to the last subcontractor on the job.
My concern is that some clients may start to see the model, which has so much more precise and digested information, as a commodity.
HOFMEISTER That’s ironic, isn’t it? We set this whole thing up to deliver better information to our teammates so we can collaborate better, and now there is a risk it could get used to turn you into a commodity. And that’s really not going in the right direction.
DUSSAULT It’s not going in the right direction, but I think we’re entering a period of transition, when things are going to sort themselves out. Clients who try to buy a model as a commodity are going to find problems with quality and all the issues that come with that. We need to convince owners that buying early is not a bad thing – even if it is a bit more complicated because the product is not 100 percent defined. I hope that in the future we will always set up the whole construction team early in the design. Then everybody will have the full benefit of model delivery.
PAULL I think model-based delivery will save designers money, especially on steel. If you don’t do a model, you get a lot of questions during shop drawings, multiple shop drawing revisions, and changes contractors want to make that you can’t easily vet. I’ve got to believe all of that costs engineering companies a fortune. It’s by far the largest number of shop drawings from any trade. If that process gets streamlined or starts earlier with a model, even if there’s more up-front cost, it has to save money in the end, because you’re always behind the eight ball. Drawings are coming in 200 or 300 at a time and we need them returned in a week, so you’ve got to build up your staff for short periods of time. That’s got to be very inefficient.
HOFMEISTER Well, it does cost quite a bit to get the model built. But as the tools get better, I think what you’re saying gets closer and closer to the truth. When we started this 10 or 11 years ago, we still had to have a few different models and the deliverables were unclear. There was the question of what governed, models or paper, so there was still a lot of inefficiency there. But the tools and the processes are much better now, and they continue to improve. What Allan said about the shop-drawing process is absolutely true. I mentioned earlier a model-based delivery project that had 97 percent come through just once, with very limited mark-ups. That’s the ideal result, because it costs us money every time a set of shop drawings lands on our desk for review.
PAULL And you don’t get any more money for it.
HOFMEISTER Right. So there is a potential benefit for our profession. I don’t think that we have witnessed the full benefit yet. But I think the owner has seen much of that benefit already, in the schedule and cost predictability we talked about earlier.
From left: Allan Paull, Steve Hofmeister, Frank Falciani. (Danny Bright photo.)
FALCIANI The enemy of general contractors or CMs is building something twice, because the schedule immediately takes a hit and then you’re scrambling; sometimes it takes months to recover. Then the profit margins of the subcontractors and the CM take a hit as well.
PAULL We don’t need the practice.
FALCIANI No, we don’t! I see us taking your model and the architectural models and exploding them up into different pieces, and getting our superintendents more sophisticated in manipulating them on their tablets or other devices out in the field. Then they can walk through a project and understand more about the designer’s real intent for a particular detail or work area. We have begun to provide Wi-Fi coverage across our construction sites so field teams can take advantage of “cloud” technology.
PAULL It’s interesting. The World Trade Center Vehicular Security Center project had very complex geometry – we used Thornton Tomasetti to do the full detailing there – and the steel contractor walked through with iPads with the steel model on them, to understand what went where and in what order. We didn’t ask them to do it. I thought that was remarkable.
HOFMEISTER I had my very first “aha” moment about 15 years ago. Structural steel detailers were building models to create shop drawings, and I happened to be in a job trailer when the steel contractor was planning an erection sequence for some tricky long-span roof work. They had a series of 3D model view extractions up, and that’s what the raising gang from the steel erector was using – not the shop drawings. It just hit me: why don’t we just actually do this with the model? Well, we didn’t have the technology then. Those 84-inch touch screens and iPads hadn’t been invented yet.
Serge has a very complicated project on his iPad right now. The erection foreman is able to zoom in and figure out which way he needs to swing something. He uses it for bolt insertion. The people with the hard hats are using this tool now.
FALCIANI We have that on our projects, too. Our superintendents and engineers have their tablets with them and they’re able to literally walk through a complex project while they “walk through” the architectural vision and the engineering requirements. It’s an amazing tool.
Will 2D drawings go away completely?
PAULL I think you will see them go. Just look at your own office. How much mail do you get? How many drawings? Almost everything I get now comes in as PDFs. I’d rather have something in PDF because then I can blow it up, I can do things with it. I think in 10 years you won’t see any paper at all. It may be five years. Market conditions will drive it. We used to spend a lot of money on blueprinting, but not anymore. As clients look to reduce general conditions costs, there’s no need to spend hundreds of thousands of dollars printing drawings anymore.
HOFMEISTER I think Allan’s 10 years is probably closer to the mark than five, and I’ll tell you why. It’s because I’m 55 years old and in 10 years I’ll be retired. My generation is much more tactile. We grew up touching paper, picking it up and putting several versions up on a wall to visualize a design. The generation that’s following us grew up with monitors. They’re much more adept at working virtually than my generation.
Technology is supporting it more and more, too. When we started delivering models 11 years ago, I worked on a small, square screen. Now everyone has two or even three big screens.
DUSSAULT I think we’re getting close, depending on the trade. We used to have phenomenal printing capacity in our drafting offices, but now several of them are paperless. They work from electronic information only. The multiple sets of drawings we used to print are gone. But the structural steel trade has always been ahead of the game in electronic information and modeling. It’s going to be very difficult for some of the MEP trades, so it will take more time. So I believe it will be more than five years before we see it across the board.
HOFMEISTER Just five years ago, a construction site would have a trailer devoted entirely to layout boards and rack after rack of hanging drawings. Now that trailer just has five 84-inch monitors – smart screens, so you can mark things up in Bluebeam, take snapshots and send them to people. We’re moving pretty quickly.
DUSSAULT We are, but like you said, there’s still a generation that may not be ready. We have one project right now where there were very few drawings in the field until they called an “old guy” out of retirement. He needed a printer and a lot of drawings. Now there are drawings pinned all over the walls of the trailer, because that’s what he’s used to working with and that’s what he recognizes.
FALCIANI I also think it’s probably 10 years out. But it is changing. I’ve worked with one client for 23 years, and my general conditions for drawing reproduction is half what it used to be and they want to slash it even more. They’re willing to pay us for iPads, but not for blueprints. We used to have a big budget for shop drawing preparation, but now everybody uses Bluebeam to mark things up.
HOFMEISTER At first we were just replacing stacks of drawings with a PDF file. But now shop drawing files can be embedded into the models. You review the live model and make marks directly into it, so that review is right there.
PAULL And you don’t lose drawings, you’re not hiring staff to fold up drawings and put them into files and record them so you can find them later.
HOFMEISTER We already have less paper, and the paper we do have has a very short life span. We print to scribble and think and doodle, and then it’s gone. Those thoughts go into the model.
Will the shift to model-based delivery change the buildings themselves?
PAULL You’ve already seen it. Nothing is square anymore. The days of a square or rectangular box going up seem to be past. Today all the buildings twist, they slope, they turn, they have holes in them…all sorts of stuff.
FALCIANI Exactly. The sky’s the limit on the architect’s imagination; they can just turn it loose. There’s a computer program somewhere that can interpret that vision, and then they turn it over to the sophisticated skylight manufacturers or to the structural steel contractors and say: “Make this all work. And make sure all your models line up with everything and all the working points are the same.” If an architect can envision it, we can build it.
HOFMEISTER I’m old enough to remember when having curves in a project, in elevation, plan or both, required you to define them mathematically. With the innovation first of CAD, then of BIM and now model-based delivery, we are able to envision and deliver much more complicated buildings.
DUSSAULT In much less time.
HOFMEISTER Yes, in the same time or less than the rectangular buildings used to take. That’s the real impact. We can digitally imagine and deliver complex schemes with no more issues and challenges than the standard buildings of the past. And it all feeds down to your shop floor.
DUSSAULT Executing that complex geometry with our current tools is not much more difficult than it was to do the simpler geometry 20 years ago with the tools we had then.
HOFMEISTER And the process of review – we still call it “shop drawing review” – that process is really changing. We hand you a model and you hand us back a model for in-model review. The process is just much more streamlined; it works so much better.
DUSSAULT Steve, do you remember on Yankee Stadium, the very first job where we exchanged models? The general contractor, owner’s representative, architect, you (as structural engineer) and me (as the structural fabricator), we were trying to figure out how to exchange those models. Everyone was nervous about reliance on the model and worried about liability, so the attorneys got involved. We started with a two-page document and everyone got that reviewed and brought it back with comments. When that document grew to eight pages, we all realized that we would never get to the end of it. So we all stopped and made the choice to collaborate in good faith, to trust each other. But not everyone can do that! Now these legal barriers are falling. It’s much easier to exchange models between companies.
HOFMEISTER There was a lot of fear then, but now we have a lot more comfort with delivering models.
DUSSAULT Exactly. Once the engineers have that comfort, the word spreads quickly. It lets architects go wild with their imaginations. So it’s going to allow us to make much more complex structures in the same construction time we used to need for a regular building.
PAULL Because the machines are computer-driven, you can make whatever members are needed. The only restriction now is how much money the client wants to spend. If they want to do something really monumental, they’re still going to have a monumental price.
FALCIANI Steve, you and I are looking at a job right now in Los Angeles in the $2.6 billion range. And without modeling, it could never be built in the time the client wants.
HOFMEISTER Because no two pieces of it are the same!
Does concrete modeling have the same benefits as steel?
HOFMEISTER I’m sure Frank and Allan have very strong opinions. I have some strong opinions, too. We have the capability, just as we do in steel, but we don’t have the partners to hand off that intelligence to like we do in the steel community. Allan gave us one example of when it works: the congestion at the bottom of One World Trade Center. But having a concrete team member sit at the table with a model and work collaboratively? We don’t see that on a regular basis yet. We were close to it with the precast concrete contractor on MetLife Stadium, a project Frank and I did together. We tied radio-frequency ID devices to the models. In my opinion, precast is more advanced technologically than the cast-in-place trade.
PAULL There are two important issues with concrete. One is where it is: a dimensional model showing the location of the edge of a slab, the face of a column or the face of a wall could be done relatively easily. The other is the rebar: it’s a bit of an art form. It can be modeled – we are starting to see more electronic production, and being able to adjust it on the fly is helpful – but will the market want to pay for that? We’re already producing concrete jobs very efficiently. A very complex building would benefit, but buildings with flat plate and sheer walls and are pretty simple to build.
I think we will see the end of people hand-drawing formwork or rebar. They will do it in some type of 3D environment instead, because again, the people coming up are not like us, who are the last of the hand drawers and are used to working on paper.
HOFMEISTER I think you hit on a key distinction. For the acres and acres of flat plate, there’s not much need for modeling. The advantage is there for situations like Allan mentioned, at the base of One World Trade Center, with the congestion, concrete meeting steel, the embeds, all that complexity.
FALCIANI I can’t disagree with anything you’ve said. It may make sense to model some highly architectural formwork, or you may need concrete modeling in a high seismic zone. But it’s not as cost-efficient as steel modeling. Plus, if steel contractors are at a seven or eight in this process, the concrete subcontractors I deal with are probably at a two or maybe a three. They just don’t have a solid reason to build using those capabilities.
PAULL I think you’ll see a change with the shop drawings. As the generations change, they’ll be drawn in some type of electronic environment as a matter of course.
What will a construction site look like in 20 years? Will there be people? Or will we have robots following instructions from a model?
PAULL I certainly hope not!
HOFMEISTER We’re already seeing some industrialized, mass-produced construction. But everyone in this room works primarily on buildings that are one-offs. With repetitive buildings, automation is more likely, especially as the cost of building robotics infrastructure goes down and where there’s enough repetition to amortize the cost of the initial set-up.
But there are other ways it comes into play. We already have human-driven robots doing scanning that feeds into a model we use to design modifications to existing buildings. So we’re on the way. I agree that it’s uncomfortable to think it could all be done by robots, but 20 years ago no one could deliver a 3D model of a building, either. I’m not a futurist, but I do see potential.
PAULL I see opportunity for prefabrication and preassembly. With curtain wall, we used to do things by stick. Now it’s all coming in module. So when we start to look at things, like bathrooms, that can come in in packages and skid across the floors, I think we’ll see more of that.
Automating heavy welds, for example, we’ve already seen that. Who wants to sit up there doing a 4-inch weld when you can set up a machine that does it a lot faster? There’s still a welder running the machine, but it goes down faster and with fewer issues.
But I don’t see someone coming in and pushing a button that causes a team of robots to go out and build the building, not when each building is unique. I also don’t see a lot of prefabrication of whole units, unless you can get into a factory-type setting and port them out to the field. The industry has tried. Remember lift-slabs? You build them all on the ground and jack them up, but there were major problems. It’s very hard logistically; there are big loads, big issues.
FALCIANI Robots are not going to build one-off buildings anytime soon, but someone’s going to devise something to do the welds in the field. Or we’ll modularize more components – particularly in MEP – so they can be brought in and put right into place. The owners want buildings faster and for less money. That is going to drive the process of being innovative.
Any final thoughts?
HOFMEISTER I think these discussions are important to have among as many people in the industry as we can. When we have conversations like this, it helps us see where the industry is going – and where we want it to go. I think integrated processes, like advanced modeling and model-based delivery, are heading up along an exponential curve. I think we’re just at the beginning and it’s going to really take off.
The post Delivering the Model appeared first on Thornton Tomasetti.