During the most recent AGC BIM Forum conference, held two weeks ago in La Quinta, CA, "trust" was a returning theme during the various presentations and discussions that I attended. Probably "trust", as in "the ability to count on each stakeholder's work" can be considered the most important constraint for IPD projects to be successful.

Because in IPD project risk and profit are shared amongst the stakeholders, participants need to rely heavily on the quality and timely delivery of each other's work.

Where contracts and claims take care of reaching a certain ‘comfort level' in the more traditionally organized projects, craftsmanship and good communication are critical in IPD to let managers sleep calmly.

"We caught ourselves calculating the ETC (Estimate To Complete) more frequently", as John Tocci stated during his presentation about the new Autodesk office project, currently built using an IPD contract, expressing the need for continuous insight in the project's status.

Craftsmanship can never replace by better information technology in construction. However, with BIM technology such as Virtual Construction teams can provide all stakeholders with the same view on the project and derived constructability, cost and schedule information, which greatly enhances the communicative capabilities of project teams, as exemplified by presentations such as the valuable Weitz/SCI contribution to the conference.

During the software forum panel discussion, an important topic was ‘change management', where we talked about how to support the need for ‘bite size' and role specific project information. With building contractors turning into ‘change managers', it is extremely important to not only understand the current status of the project, but also to understand which decisions got the team there.

During the development of the upcoming "Vico Office" software, we are continuously keep this in mind - our goal is to provide (IPD) project teams with tools that helps them to focus on craftsmanship, not on explaining project data!

The central question for this post is: how do you plan the work on a ‘Virtual Building’; a project almost as complex as the ‘real thing’, with contributions by many participants, representing all involved building disciplines?

As an initial answer to this question, I will give three examples of companies who have found a solution for this by implementing concepts that allow for planning the ‘richness’ of the information captured in the VDC models they build.

A few weeks ago, I received an e-mail from Brian Lewis of ACLA, an architectural and urban planning company who are currently working on the realization of a large new campus. ACLA are long time users of Vico Software and have used the software for design, cost estimating and scheduling purposes.

In response to my previous blog about the Content Plan, Brian shared with me a plan, in the form of a spreadsheet, that the team uses to define responsibilities and scheduled deliverables to create the VDC (Virtual Design and Construction) model for the project. By using color codes (red, yellow, green), the status of the work to be performed by each of the team members can easily be identified. Capturing the work to be done this way helps the team to check that the model contains the required information for decision making at the right moment, and makes participants aware of the interdependencies between work to be completed by all involved disciplines. I thought this was an excellent example of a ‘super level’ for the Content Plan, which greatly helps to coordinate VDC work!

ACLA’s VDC Work Plan

Over the last 6 months, I worked with a team of Vico colleagues and Webcor Builders’ VDC team in San Francisco on a framework to plan model, cost estimating and scheduling information for projects. The concept resulting from this effort is the ‘Model Progression Specification’. Jim Bedrick of Webcor has written a great article, describing the concept and formalizing the used metrics for AEC Bytes.

The MPS consists of a set of formalized ‘Levels of Detail’, labeled with a number, and together with a typical use of the VDC information at that LoD. After defining the LoD’s, Webcor can now define the ‘progression’ of the VDC model throughout the project stages. The progression is defined per Uniformat ‘System’ and allows for specifying the goals for ‘stages’ in the project. Important to note is that design stages can now be released in a higher frequency, because only parts of the project have to go to a higher LoD, not the whole project at once! Working in stages means: more frequent feedback, which results in better opportunities to manage the project, which again allows for making changes when it is still possible. Webcor is now building the content (including model, cost and schedule information), that will allow for building VDC models for all defined LoD’s.  To learn more about the model progression specification at Webcor, please be sure to view the archived Fridays with Vico webinar on the topic.

In a third example, Vico is currently working with St. Joseph Health System (SJHS) in Southern California to define an outline and content for a BIM-driven process for cost estimating and scheduling for a healthcare owner organization. Again, the work includes defining standard LoD’s and describing typical uses for each of the LoD’s. With the defined LoD’s, a standard Model Progression Specification will be defined for SJHS. The MPS will also define when during the project SJHS’s partners are getting involved. By working with these partners to build VDC content, SJHS is defining a streamlined Model Progression process with standardized LoD definitions for input from contractors, designers and consultants. By implementing the MPS driven process, SJHS’s goal is to use VDC based information for better and more timely design coordination, cost and schedule feedback, which will provide better input for decisions (‘go’ / ‘no go’) throughout the design and preconstruction phases.

Learn more about other work going on at St. Joseph Health System, including a virtual mock up project by viewing the archived Fridays with Vico webinar on the topic.

Three examples of three different types of companies – an architectural firm, a general contractor and an owner organization – all making the next step in the implementation of VDC technologies and concepts.

I am sure that many companies who are implementing BIM and VDC are working on similar efforts – I would be very interested in learning about other approaches and experiences. Give your feedback to the concepts described above to share your thoughts on ‘how to plan a virtual building’!

September 6, 2008 

When your organization has decided to start Virtual Design and Construction (VDC), you will be confronted - or you have already experienced - the need to plan VDC content to achieve the best model-based estimating and scheduling results.

In my previous post, I wrote about Vico's Content Plan concept: with the Content Plan, you define all elements and their requirements for modeling, estimating and scheduling, before creating the building model.

Defining a Content Plan is a team effort, and by thinking about the information that should be included in the VDC model with the project team, the first benefit of creating integrated project information is already obtained: all team members involved have to define which information they need in order to make the decisions that they need to make in the current project phase. By doing this before everyone starts working on this information, a shared ‘To Do' list for the team is defined and requirements are communicated!

The information, demanded by team members, can vary from ‘visual study of the façade' to ‘the net volume of concrete on the 3^rd floor'; each requirement results in a level of detail for a part of the project, for example per ‘Uniformat' category.

The proactive approach of information management with a Content Plan helps the project team to really focus on those parts of the project that matter most for the current phase of the project. Those parts of the project that require detailed decisions or value engineering, for example ‘the concrete structure in the lower floors of the building', should be included in the model with a higher level of detail then, say, ‘floor finishes', because detailed decisions don't have to be made for those, yet.

By capturing the requirements per project phase using a content plan, the team can get a better understanding what information is needed (the concrete structure design needs to be finished) and which information has lower priority (basing the cost of floor finishes on an average cost per square foot is acceptable).

Formalizing the content per phase also allows for defining the actions, required from each team member - more about that in my next post.

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.

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.

Vico is working on a number of initiatves with our Partners to assure more streamlined processes.  First, we have built Vico Office as a BIM-neutral platform.  This allows users to publish BIM files from Revit, Tekla, and ArchiCAD (no re-modeling, no waste).  Secondly, working with Trimble's LM 80 and Total Station, users can plot points inside the BIM which can then be placed out in the field.  Conversely, points in the field can also be marked and overlayed on the BIM for a comprehensive quality check.  Thirdly, CMiC and Vico spread the information contained in the 5D BIM to the various departments throughout the contruction firm, keeping the bottleneck of change management under control.  Finally, Vico Control allows superintendents to walk the jobsite and note the percentage complete, allowing immediate "action" in the PDCA cycle. 

Companies that successfully implement systems that allow for deriving project status instead of contriving it, will be prepared for the future.

What is your firm doing to institute lean construction methodologies?  To institute BIM or other technologies to make production control and "checking" more streamlined?

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.