Meet the manager: HOK's corporate CAD director

30 Apr, 2001 By: Cadalyst Staff

I first met Mario Guttman about a decade ago, when both of us were finding our way in the architecture and engineering corner of the CAD world. We showed up at the same San Francisco Auto-CAD user's group meetings and bumped into each other at various design firms and software companies where Mario worked and I consulted. Each time, I was struck by Mario's combination of down-to-earth good sense and optimistic imagination about where CAD could go in design firms. It didn't hurt any that we share a delight in skewering the inflated rhetoric of software hucksters and industry pundits.

Lots of people breathlessly proclaim their "vision" of CAD, but with very little grounding in reality. Others have the hard-won experience to know what works and the skeptical frame of mind to deflate pretentious, technology-fueled fantasies. But these skeptics often lack a sense of possibility and a willingness to take risks. Mario is one of the few people I've met who successfully straddles this divide. How does he do it, and what can we learn from him?

Anatomy of a mega-firm

Since 1998, Mario has been corporate CAD director for HOK (Hellmuth, Obata + Kassabaum) Group, Inc., the world's largest architectural firm. HOK has 25 offices world-wide with more than 2,000 employees, about half of whom use CAD. As part of the Corporate Advanced Technology Group, Mario reports to the chief information officer and chief knowledge officer and attempts to ride herd on 25 CAD managers who are responsible for CAD procedures in the individual offices. Mario coordinates firm-wide research and development activities and helps the firm formulate its long-term CAD policy.

Mario Guttman"Our company is a microcosm of the industry as a whole," Mario points out. "Although the company is big, in some ways we act like 25 smaller offices." The size of the company enables HOK to work on immense projects, but the "small office" character of the individual offices enables greater creativity. According to Mario, the individual offices do a lot of useful things with CAD. Part of his job is to "harvest that richness."

At the project level, each project is assigned a data leader who coordinates the team and the CAD manager. This person is responsible for all of the information technology issues, including CAD, databases, and other software.

Communication is a big issue for a firm of HOK's size (and not just for CAD management—the firm shares many projects among the offices). The CAD managers and Mario meet via conference call once per month. They use Net-Meeting to explore problems and teach solutions. Sometimes they even collaborate on documents this way.

CAD infrastructure
HOK defines three tiers of computer technology that must be managed: the computers are on the bottom tier, the networks are in the middle, and the computer applications and their proper use are on the top tier. When I first met with Mario to discuss his work as HOK corporate CAD director back in September 1999, the bottom two tiers consumed most of the information technology staff's energy, and there was little left over to focus on software applications.

"We know it's hard to use CAD software," says Mario. "We think we could use it much better than we do. So much of our effort is drained off into computer problems (hardware and network issues). Our corporate-wide approach is to provide good solutions to these problems so that people can focus on the top tier issues of how to do better design and automate our processes."

A year and a half later, HOK has made good progress in reducing the bottom-tier requirements by defining standardized configurations and outsourcing and leasing hardware. In the middle tier, it has moved to a worldwide VPN (virtual private network) that uses the Internet as a network backbone. This network is fully managed, which means that a service provider supplies and supports all of the hardware. His network also is fully meshed, which means that every office has a direct connection to every other office. At the top application level, HOK negotiated a firm-wide enterprise agreement with Autodesk to provide Architectural Desktop and 3D Studio VIZ to all users on a maintenance plan. "We've done a really good job of securing the infras tructure, but we're still struggling with how to really use the software," he says. "We chose Architectural Desktop because we think that the object model will provide a solid basis for the future, but we are engaged in an aggressive customization effort to make it really work for us."

The changing CAD landscape

"There are two things we can do with CAD," Mario notes. "First, we can figure out how to improve the things we're doing now so that we do them faster, better, and cheaper. Second, we can try to add value— that is, do more, do it better, and be in a position to charge for the added value." Because Mario came to HOK with a CAFM (computer-aided facilities management) background, he was especially attuned to the second approach and is one of its strongest advocates at HOK. "What's really great is when you find something that you can get for free because it turns out the data is already present somewhere else," he says. "Our clients are beginning to see this kind of added value. We want to be there to deliver it so that they don't look outside our profession for a solution.

"We're trying to reassert the meaning of ‘CAD manager' as one who is responsible for the office's use of computers in the aid of design. This is a new kind of employee, someone who is a technical specialist within the design professions. They need to be able to apply software skills to high-level problems, such as defining an Architectural Desktop curtain wall object that also meets architectural realities. The exciting thing is that these people are becoming really important to the firm. We are beginning to see a real career future for them that never used to be available to the people who got all excited about the speed of their processor."

I remember another user's group meeting where Mario and I bumped into each other. The group was discussing dumb things that AutoCAD users do and how to prevent them. Others suggested various Draconian ways of crippling AutoCAD in order to lock users out of potentially dangerous actions (disable menu options, lock down files, etc.). After several minutes, Mario spoke up: "Actually, I think that we should just hire really smart people and pay them a lot of money." Mario's remark deflated in the gentlest way possible the fantasy that making computers "smarter" makes up for having unskilled people using them.

Not that Mario underestimates the importance of supporting and managing those smart, skilled people: "HOK is very visionary and seeks to distinguish itself by being a technology leader. However, we are very cautious about making radical changes in our ongoing processes. Our implementation of Architectural Desktop, for example, is based on carefully staged steps.

"First, we are getting teams comfortable with the AutoCAD 2000 features, such as the plotting changes. We're then introducing the use of a few selected objects, such as doors and walls, so that users become comfortable with proxy objects, etc. Finally, they move on to more advanced features such as automatic sections and schedules. Throughout all of this, we are very conscious about not disrupting the ongoing workflow."

Historically, HOK has made a high-level commitment to software objects. The firm has made a major commitment to the IAI (Industry Alliance for Interoperability) and plans to prototype projects using the IFC (Industry Foundation Classes) this year. At the same time, Mario is concerned that the industry hasn't really come to terms with what this all means: "There isn't a shared understanding of what objects really are or how object models will really work, and there is a tendency to gloss over these issues. It turns out that objects are a real pain in the neck sometimes. Either they are so dumb that you end up designing ‘ranch-burger' buildings, or they are so complicated that you need a Ph.D. to figure out how to use them. Most of our design is really concerned with the subtle variation in building components, and we're not always all that interested in new ways to create the boring stuff."

Training and CAD standards

One of Mario's jobs is to develop policies and resources for training. "Architects have a funny attitude about CAD training," he says. "Nobody says we should send the team out for detailing training. HOK has an aggressive education program, but it's focused more on our collective process improvement rather than traditional training. This might entail a work session, documenting a procedure, or any number of other techniques. Most of the real learning occurs on the job where three key components are present: an issue that needs to be addressed, a real solution that is available, and an opportunity for the user to exercise the solution.

"Many times the blame for a problem project is assigned to lack of CAD training because it is a convenient scapegoat. But sending people out to classes is very expensive and doesn't improve performance all that much. The real problem is figuring out what to teach. There's no point in learning all the options of the Pedit command. What people need to know is how to put the project together, and unless we have a really good answer, we're just wasting their time. When we do have a good solution, it spreads by itself.

"We've also found cultural barriers to training," he adds. "People are not comfortable sitting at their desk and doing an exercise, because they feel pressure to be working. We have also found that they are often afraid to admit their ignorance of something. This is ironic, because HOK requires a minimum level of continuing education and pays for much of the time and tuition costs.

"Frankly, I don't have much patience for people who aren't learning technology. The world is changing. You need to take your own time to go out and learn a whole lot about a whole lot of things. You can't wait for others to spoon-feed it to you."

Mario is equally outspoken about CAD standards: "I'm really bored with CAD standards. It doesn't surprise me that architects resist when some martinet gives them grief about a layer name. Our CAD standards represent the failure of the software to adequately serve the process we are automating. We shouldn't even have layers."

Curmudgeon vs. optimist

Mario ended our interview with several thoughtful gripes about the current state of CAD software development: "It really irks me when people refer to CAD as ‘just drafting.' Drafting is a very subtle process of abstraction that has a long history and culture attached to it. It's arrogant if software developers think they can redefine the construction process just because they know how to program computers. I've worked on both sides of the fence, and architecture is a lot more mature of an industry than software development.

"One of the myths that has come out of this dismissal of drafting is the notion that we want to model the whole building and then automatically extract plan, section, and elevation views. This is an interesting theoretical concept, but it obscures subtle issues about how information is represented in construction. In many cases, we don't want to define the building that specifically, and even when we do, it is an immense task. We work hard to create a process that is based on a shared object model because we see the model as adding value in a collaborative environment, but I don't have any illusions that it will be easier.

"At the end of the day, the real issue with CAD software is that it isn't very good. It's frustrating that we still deal with issues that we were dealing with years ago," he says. "The software industry seems to be collecting an inordinate amount of information, through surveys, user groups, etc., without really getting it. It's very hard to find people with true design experience within the software community. That's why we're so excited about the recent hiring of Phil Bernstein as Autodesk's vice-president of the AEC market group. He hails from Cesar Pelli's office and was a Yale professor. He brings some ‘adult supervision' to that part of the company. We see Autodesk doing some progressive things in its StudioDesk project, which is well-staffed with seasoned architects. So, even though I might sound like a curmudgeon, I'm pretty optimistic. And I'm having a really good time."

About the Author: Cadalyst Staff

Cadalyst Staff