August 18, 2008 

While the use of Building Information Models is getting more and more popular, it seems that the full advantage of using the model-based approach for informed decision making when it comes to cost estimating has yet to be experienced. In my opinion, the reason for this, is that many building models are created with a ‘hit-and-run' approach: Building Information Models are created for projects without considering what the desired output should be.

The models that are created this way can be great for visualization and maybe even for coordination. But, because there is no defined set of requirements before the modeling work takes off, may be useless for estimating and scheduling...

Why is that?

It all comes down to quantities. It's the quantities that drive the estimate and the (integrated) estimate and schedule and when some of the quantities that are needed are not available, the ‘5D Mission' fails: only some of the quantities extracted from the model are useful. Some quantities are too detailed for the stage of the project, other quantities are simply missing. The only use for the model for estimating purposes in this case is to find opportunities to find model-based quantities that happen to be useful. I call this the ‘reactive approach': trying to find some good use for the model that was created when it's done.

There's a simple remedy for missing the opportunity to extract quantities from models that do match the requirements at a certain stage in your project. At Vico, we call that remedy the Content Plan. The Content Plan is nothing else than a document that captures the information needs of all of the members of the project team for the deliverable at a certain stage in the project. When the project team gets together to define these goals before the model gets built, it is much easier to build something that fits and really turns into a tool that helps your team make those important decisions... Stay tuned for more on this topic!

March 25, 2008 

Recently, I attended a teleconference organized by the Association for the Advancement of Cost Engineering (AACE). The purpose of this call was to kick-off an effort of the AACE's BIM Technical Committee to define requirements and standards for the use of Building Information Modeling for quantity take-off and estimating. Among the participants were owners, contractors and universities, all eager to contribute to this project.

The start of this project is an important and highly interesting development in the adoption process of BIM in our industry. However it is evident that Building Information Models implicitly contain a large collection of project information, which is captured by means of CAD objects, a lot of this embedded information is not directly available for users to, for example, create model-based estimates and schedules. Clearly defining the real needs from BIM for estimating (and scheduling) will therefore proof to be crucial for a successful adoption of BIM for construction.

Of the set of model-based information that is directly available for BIM users, now even through (semi)direct connections to estimating software packages, only a sub selection is useful for estimating. Regretfully, the Level of Detail (LOD) of this information (typically existing of square footages, numbers of elements and volumes) makes that the usefulness of the extracted BIM data is basically limited to estimates in the conceptual and early design phases.

Of course, the question "what can be expected from a model-based estimate?" arises.

The answer is: it depends on the LOD of the model-based quantity take-off and how well requirements are documented.
In the projects that we at Vico have worked on over the years, we have identified a set of quantities and element attributes that are required for detailed estimates and schedules that go beyond the conceptual and early design phases.

Vico products derive the available information from the Building Information Model and then process it in such a way that cost and duration of construction activities such as formwork, finishes and MEP, which require more specific quantity take-off input with a higher LOD, can be based on that model. The set of integrated model data that is created the Vico way can be used from conceptual phase to construction, making BIM possible throughout all project phases for estimating and scheduling.

January 10, 2008 

In 1986, Dr. W. Edward Demming, author of the infamous book "The Toyota Way", which has become one of the great inspirations for our team at Vico, defined a model that applies to virtually every imaginable process. It is often referred to as the "Demming Wheel" and consists of a circle divided into four equal parts, each one representing a typical phase for a management process: Plan, Do, Check and Act.

People who work in construction management jobs go through this cycle numerous times per day. They "plan" what needs to be done next, "do" the required things to make that happen (attending meetings, placing phone calls, sending e-mails), then "check" if everything is going according to plan and finally "act" if something is not going according to plan, after which the cycle restarts.

It is typically the third phase in the PDCA cycle - the "checking" part - that is taking a disproportional amount of time from construction managers. The vast amount of both paper based and digital, but foremost disconnected information that provides them with project status updates requires a lot of processing time.

It is common sense that the human brain is limited in its capabilities to process large amounts of data and to subsequently derive an objective view on the actual status of a project. It is therefore, that jobs in construction management require many years of experience and exceptional skill sets in order to make sure that projects are completed on time and with the desired quality by interpreting available information the right way. The majority of people that are currently occupying these crucial jobs in the construction world are from the "baby boom" generation, which is expected to retire over the next 10 years.

The challenge that the construction industry is facing today is to make sure that the "C" in Demming's cycle is going to be performed successfully in the future as well. That is: after the people with the exceptional skill sets and years of experience that are running projects today will have left the site. Automation and smart reuse of knowledge seems to be the horse to bet on. Companies that successfully implement systems that allow for deriving project status instead of contriving it, will be prepared for the future.

November 20, 2007

When you know you need to be flexible and that your project is due to go through lots of changes, limit the level of detail and/or complexity of your model: use "easy-to-update" geometry like Zones and Walls and avoid highly detailed geometry like façade structures unless you have to build a geometrical model for visualization purposes. Simply use numerical model components to compensate for any missing geometry. The combination of key figures and simple geometry already provides you with excellent input for an estimate. With specific early phase calculation formulas (formulas are called "Recipes" in Vico's estimating solution), you can generate detailed estimates with limited input.

During the bid phase, when detailed design information is available and when most decisions have been made, you can add more complex geometry to your model. However, make sure that you know for which purpose you are doing this: define the quantities and/or properties that you want to extract upfront, so you do not spend time defining that curtain wall façade in a high level of accuracy: sometimes a representation by means of a simple wall already gives you with the quantity you need! (For example: surface area of the façade.)

Along with the design and specification documents of your project, the level of detail of your estimates will evolve over time. With model based estimating technology, you can simply replace the numerical or abstract geometry based quantity input from the early phases with more detailed quantity data when it becomes available. With tools like Vico's "Cost Manager", you can even compare previous versions to current versions and perform target costing.

Once you have experienced the power that model based estimating gives you (again: this can be either a geometric or a numerical model), you will find that you can be much more flexible and responsive in your work as an estimator.

Having to spend significantly less time on quantity take off (when you go through iterations of alternatives, you can simply adjust your model's parameters) you can spend more time on what is most important in your job: managing cost and selecting the best solutions for the project, really using your knowledge of construction cost. Answering the "why?" questions during meetings will be easier: you know which model parameter was used as input for your cost calculations when changes to the design were made. With geometrical cost calculation input, you can even visualize where cost is coming from.

If you are convinced that model based estimating makes sense, but you don't know where to start, just pick a project and start by enlisting the types of quantity data you typically use as input for your estimates. Then, take the estimating formulas and knowledge data that you already have and copy them to a model based estimating application. Based on the required quantity data, build a simple 3D model - you will see that a lot of the quantities you need can be extracted from simple model elements like "Zones" and "Walls". For quantities that are missing, you can just define them as a numerical value - just replace them with geometry based data when you're ready for it. In no-time, you will have created your first model based estimate.

By doing your first model based estimate this way, you will get a feel for how it works; even when the only thing you're extracting from a geometrical model is "floor surface area". Also, you will start building your own company knowledge database, which you can re-use in your next project. You will experience that with every project, the amount of information that you can generate from models of limited complexity is increasing rapidly.

Over time, you will create templates for typical geometrical (for example typical wall and floor slab structures) and numerical (for example typical project key figures) models, which will provide you and your company with standard model based estimating guidelines. Predefined formulas ("Recipes") will make it easier for new colleagues to create accurate cost estimates, based on your company's knowledge database.

October 25, 2007

In this blog, which will often be produced while being airborne, I will write about the new approach to estimating that we at Vico have been practicing for about 3 years now. It’s about a technology that has been around for much longer but has become more and more popular with the increasing adoption of Building Information Modeling. This blog is about Model Based Estimating.

One of the goals of this blog is to become a platform for people who are interested in or practice model based estimating. If you’d like to share your experiences with and knowledge of the concept, feel free to send an e-mail to "flying.dutchman@vicosoftware.com”

As a start, I want to stretch the term “Model Based Estimating”. When talking about Model Based Estimating, it is automatically assumed that it involves the use of 3D geometric building models. Although 3D building models do indeed provide great input for estimates with their ability for automated, accurate quantity take-off, model based estimates can also be based on more abstract, numerical, models (which is often referred to as “key figures”, defined in the conceptual phase of a project).

I will use the term “model” with the description that can be found in most dictionaries: “a hypothetical description of a complex entity or process”. In construction, we create such descriptions on a day-to-day basis: making abstract representations helps us to make decisions on complex situations, like the estimated cost or the budget of a project.

Both types of models (the geometric building model and the numerical model consisting of project characteristics), or also combinations of both, give estimators more decision making power. By deploying the concept, you can actually “model your estimate” by developing alternatives using models to define variations. Analyzing alternatives is easier with models because of their flexible nature; going through iterations makes it easier to decide on the best solution.

Over the past 6 years, I have been working with the concept of model based estimating. I started using the technology as an engineer at the preconstruction engineering department of a construction company and after that, I continued as member of the Graphisoft – now Vico – “Virtual Construction” team. During this timeframe, I have been involved in numerous projects, a lot of them so called “pilot projects”, in which model based estimating was explored. The projects I worked on were projects, in which model based estimating has been applied vary from projects in the conceptual phase, to projects in bid phase.

The amount of available information and the variation in this information throughout the stages of design to pre-construction differs significantly. However, for all phases, the same rule for successful deployment applies: define what your goal is before you start to create your model. Like in math: write down the problem statement and subsequently define the constraints for the model that will help you to solve it.