The Yale Residential Colleges as an Autodesk® Revit® Pilot Project: Planning, Training, and Staffing

The Yale Residential Colleges as an Autodesk® Revit® Pilot Project: Planning, Training, and Staffing

Melissa DelVecchio, Partner, Robert A.M. Stern Architects

William West, Associate, Robert A.M. Stern Architects, Co-Speaker

Charles Prettyman, CAD Manager, Robert A.M. Stern Architects, Co-Speaker

AB5481 Firms implementing Autodesk® Revit® face a number of challenges: What type of pilot project is best? How will the team be trained? How can a pilot project support a larger transition to BIM? In this case study, a partner from Robert A.M. Stern Architects will describe the process of implementing Revit on an unusual pilot project: the Yale Residential Colleges, a large and complex university project. She will be joined by a member of the project team and the firm's CAD manager to provide a balanced presentation from the perspectives of management, design process, and infrastructure. This class will address how senior management can work with staff to develop and apply a clear and targeted implementation plan. Topics will include techniques for training staff, selecting an initial modeling strategy, developing and testing specialized modeling techniques required by the project, and strategically building team expertise for leveraging on future projects.

Learning Objectives

At the end of this class, you will be able to:

§ Effectively plan for a transition to BIM for a large team or small office

§ Develop targeted training exercises to build varied team expertise

§ Better integrate BIM into your office's design workflow

§ Initiate testing and analysis to help inform modeling strategies

§ Implement new communication tools and effective knowledge capture and sharing

About the Speaker

Melissa DelVecchio first joined Robert A.M. Stern Architects in 1998, and has been a partner in the firm since 2008. She is the project architect for the Northwest Corner Building at Harvard Law School, two new residential colleges at Yale University, and the Stayer Center for Executive Education for the Mendoza College of Business at the University of Notre Dame. Ms. DelVecchio received her Master of Architecture from the Yale University and her Bachelor of Architecture degree from the University of Notre Dame. She has served on juries for design studios at Yale University, the Rhode Island School of Design, the University of Notre Dame, Temple University, and an awards jury for the Society of American Registered Architects. She has taught courses and planned travel programs for The Institute of Classical Architecture & Art in New York City. Ms. DelVecchio also serves on the Advisory Board of City Polytechnic, the New York City high school for architecture, engineering, and technology.

Introduction

This presentation is a case study of a work in progress -- Robert A.M. Stern Architects transition to Autodesk® Revit® using the new residential colleges at Yale University as a pilot project. What makes this case study interesting is that it runs counter to the conventional advice to "choose a small and simple project" because it makes for an easy testing ground. Based on our recent experience, we're going to argue the opposite -- that a big, complicated project with a large team can better set the stage for an office-wide BIM transition.

We did not choose our pilot project; our client chose it for us -- after we completed our Concept Design in the spring of 2009, Yale asked that we use BIM for subsequent phases. This decision coincided with the financial crisis, and in addition to switching to BIM, Yale slowed down the project schedule to allow for a longer fundraising period. Fortunately, this left us with an opportunity to plan for and execute training, and to adjust our phase schedules to better suit a BIM workflow.

The Yale Residential Colleges project is an uncommon project in many respects. The building is big -- at 500,000 square feet and about 4000 rooms, the two new residential colleges will provide a total of 850 student beds and allow for a 15% increase in Yale's student body. The building has a high surface to floor area ratio -- its 1.5 miles of non-repetitive Collegiate Gothic façade wrap around 9 courtyards, and includes 80 different window types, 4150 individual windows, and between 400 and 700 individual pieces of architectural ornament. The building is volumetrically complex -- 5,530,000 cubic feet distributes around the site in series of attached buildings ranging from two to eight stories including several towers, one of which is 190 feet tall and will contain a set of bells. The site is unusual -- a triangle with a grade change of one full story, surrounded and divided in half by intertwined Yale University and City of New Haven in-progress utility projects. The program is extremely varied -- including rooms as diverse as small student rooms, large gathering areas such as dining halls and libraries, offices, seminar rooms, faculty apartments, two large houses for the College Masters, an enclosed loading dock, kitchens and support space, and a basement bowling alley, none of which stacks easily. Lastly, our team is large -- ranging from 18-30 people during the last four years on the architectural
team alone.

We were convinced that if we planned our BIM transition using the same rigorous research-based approach we apply to our projects we could prove conventional wisdom about Autodesk® Revit® pilot projects wrong. Why couldn't bigger be better? We knew that a large, complicated project would draw us deeper into the software's intricacies, and force us to leverage the software's most powerful tools. We intuitively understood that this would allow the team to learn more, and faster, and were convinced that putting the software into the hands of a large, talented, and diverse team would lead to modeling innovation and fast-paced discovery about Autodesk® Revit® and other BIM tools.

Now that we have been working in Autodesk® Revit® for three years, we realize some of the potential pitfalls of starting with a small project. First, a smaller project allows, even encourages, idiosyncratic modeling and workarounds that will not work on larger projects, and may be painful to discover further down the road. Second, while a smaller project may seem easier, it will fail to trigger some of the most powerful BIM tools that result in the greatest efficiencies. Why saddle early adopters with all the extra work to model the building on a project where they will not see all of the benefits? Lastly, it is actually beneficial to be challenged by your modeling process. Ideally, a pilot project will be one you might not have been otherwise able to accomplish without BIM -- challenge yourself so that you and your team will learn the most.





The summary below lays out a few of the lessons we have learned to date, and spells out our current thinking about how our experiences thus far can inform a larger office-wide transition.

Effective Planning For a BIM Transition With a Large Team or a Small Office

Think strategically about your pilot project. A few questions at the outset will help you to better define training exercises and modeling strategies, and plan ahead for a future office-wide rollout.

§ How large is the project? How large is the team?

§ Is your pilot project a full services project? Are you working with another architect?

§ Will you be required to produce both 2d drawings and a BIM model? How will the Client and Contractor use the BIM model?

§ Will you be able to convince your client to implement a phase-by-phase schedule that is suited to BIM?

§ What are the existing skill sets of your staff? How might these translate to modeling skills? How might their skills inform your team structure?

§ What will be the specific modeling challenges for your pilot project - building size, spatial or technical complexity, geometry, schedule, working with a large or complex team?

§ What will happen after this pilot project? How can you train this team so that they are poised to support a larger roll-out? How can you build both broad skills and specialized expertise? How can trained staff from a pilot project can help to share knowledge and train other staff?

Developing Targeted Training Exercises

A first step in the process is to determine how exactly you will train your team. Plan this out as carefully as you would plan any project. Think beyond basic software training to how you might customize the training to the project at hand. In our case, creating an environment for low-risk experimentation led to greater innovation in modeling techniques and a culture of shared learning and experience.

§ Our office has found that training specific to a project, delivered at the outset of the project, is very effective.

§ Design your training exercises to include one exercise suited to overall skills, and one that is an opportunity for more detailed and complex work.

§ Make sure the training exercises build on one another, reinforce class material, and develop skills specific to your pilot project.

§ Find a suitable test case that, although it may be smaller and simpler than the building, encompasses the modeling issues your team will face on the project.

§ Your basic software training will take about two weeks. Holding sessions during this period that last half a day, with the remainder of the day spent working on the training exercise, will build skills slowly, improve retention, and allow time for staff to formulate good questions.

§ Don't give specific modeling guidelines for the training exercise -- allow for experimentation and you will be rewarded with innovation.

§ Don't limit exercises to modeling or construction documentation -- explore presentation tools, working with design options etc. Use the opportunity to discover how Autodesk® Revit® can support your entire workflow.

§ Plan a working session where team members present their modeling progress to one another for discussion and evaluation. Do this more than once during the training schedule.

§ Well-informed senior managers are critical to a BIM transition. Find a way for everyone to participate in the training, and consider implementing additional targeted training for partners, associates, and project managers.

Assigned Training Exercise: Week One

Training Results: Week One

Training Results: Week Two

Integrating BIM into your office's design workflow

You will not find Autodesk® Revit® to be suited to every task during every phase on every project -- it's just one new tool in your toolkit. Think carefully about how you can support your office's design process with BIM, rather than letting BIM shape your design process. For our office, physical models, hand sketches, and even 2d AutoCAD drawings remain critical components of our process as tools to very quickly investigate options. Autodesk® Revit® played an important role in supporting all of these efforts, even our creation of presentation drawings.

§ Autodesk® Revit® is a powerful database. You can use this to support your development of Construction Documents, but also as a powerful tool for creating presentations.

§ The same tools allow for useful visualizations of building components for design review - this is critically important for a large and complicated project, saving time and making your reviews more effective and accurate.

§ In combination with other tools, (Autodesk 3ds Max, M-Color, and Adobe Photoshop) we were able to leverage the massing and material information in the Revit model to support the production of elevation drawings with the qualities of a Beaux-Arts watercolor rendering.

§ Don't forget about hand sketches and quick study models… Sometimes this is the fastest way to work through design options. Use BIM to understand the problem faster and to give you an underlay on which to sketch and work it through.


Testing for Design Workflow - Rendering Elevations

Initiating testing and analysis to inform modeling strategies

Rigorous trial and testing of various solutions can have a big payoff. This software is still relatively new and everyone is still learning. Seeing modeling strategies abstractly and testing their suitability to your workflow can help your team to work more efficiently.

§ There is no right way to model a given building or element. This is both liberating and daunting.

§ Test your modeling strategies before you implement them. Different strategies will be required for different projects. Strategies that work well for one team or one building may not work on a project of a different scale or type.

§ Devise simple abstractions that will give you quantifiable data about your modeling strategies and allow you to test efficiency of both workflow and file size.

§ Decisions about your modeling strategy need to be strategic and carefully implemented.

§ Stay nimble. The best strategy in one phase might require rethinking in another.

Testing: Modeling Strategies for Dormers

Implementing New Communication Tools and Knowledge Sharing

Working in BIM is a more highly collaborative process for architects than a traditional workflow. Although there is much focus in the profession on communicating and coordinating models between the architect and their consultants, or the architect and the contractor, far less attention is focused on communication within an architect's project team.

The large team size on the Yale Residential Colleges brought this issue to the forefront, but we saw very quickly how the right communication tools could be valuable to teams of any size, and could even be beneficial to our office as a whole. Your staff will be learning very quickly during the training period -- this knowledge can inform your office standards and workflow. As a firm, you should be ready to capture and document this knowledge. We found communication and knowledge-sharing to be critical to a successful Autodesk® Revit® implementation and discovered a variety of tools that could help us to communicate effectively.