1990s Concepts
WorkPod CubeOffice and Armature system
Capital Holding Corporation - 1990
refining the navCenter concept
Virtually all my architectural work of the 1990s was directed toward the refinement and actualization of the navCenter concept and refining the tools and processes by which we design, build and operate these environments
The floor plan for Capital Holding, drawn in 1990, is the first layout showing the workPOD™, CubeOffice and Armature™ Systems [link: armature]. Elements of the Armature concept were incorporated into their navCenter built the following year. As a WorkFurniture system, the CubeOffice was prototyped in 1996, the workPOD in 1998 and the Armature in 1991, 1997, 1998, 2000 and 2002. With these, we have created an entire system covering from wall-to-wall, floor-to-ceiling. An architectural scale, flexible WorkFurniture system that provides easy user controlled adaptability, arrangement and branding options.
Capital Holding Center opened in 1991 with built in place steel Armature system, adjustable WorkWalls, ramps to different floor levels, with multimedia technology and KnowledgeWorker loft area.
At the Borgess NavCenter [link: borgess navcenter], the entire 8,000 square feet was reconfigured (along with sound, video and computer network) by a small team of Taylor KnowledgeWorkers in 3 hours. The goal was for end-users to do this with three simple tools and no “expert” or “tech” support. As a test of this, in a 1999 7 Domains Workshop, the participants walked in to an almost totally deconstructed space. They were charged with setting it up in two and a half hours from scratch and to do so in a way that it had never been arranged before. They succeeded.
Captial Holding Radiant Room
First generation POD developed by Paul Lyons installed at the Palo Alto knOwhere Store in 1997. This POD can be configured as a “S” curve with work stations on either side or as a enclosed circle forming an 11 foot “office” space. Wiring runs in a channel at the base so the entire set up can be feed from one source. Each of the sections are built up from a choice of several components and can be moved as a unit then reconnected as an assembly.
The surprise with the workPOD was that early users did not know how to use it. Originally, we built twelve and employed them in various projects. The response was divided. Half the people loved the Pod and the other half disliked it intensely (almost everyone liked the CubeOffice system). This did not surprise us but what did was that even those who liked the Pod had difficulty learning how to use it well.
Of course, I have wanted a POD since I first developed the concept so I took advantage of this situation to make one at the Palo Alto knOwhere Store my home. I was, for over three years, a working Guinea Pig [link: matt’s workspace at knOwhere] for POD use and research.
In recent years the general acceptance of the Pod has grown. It is very popular now the big issue being affordability with the Paul Lyon’s version. The new Blackburn Pod (2002) will cost about half a much. It also accomplishes our goal of rolling as a single unit. As it is built from CubeOffice components it fits well within the entire system and production schedules. The completion of the Vanderbilt project in mid 2002 actualizes the vision and intent of the 1990 sketch. A 12 year development process from idea to regular production.

In 2002, Bill Blackburn developed a Pod for the Vanderbilt Center for Better health based on the CubeOffice system. This is very close to my original Pod design which I conceived for Vanguard in 1995 yet much easier to build and configure in a wide variety of ways. The VBCH navCenter was built and installed, mostly from stock components, in less than 30 days.

Click on the graphic (left) for a history of MG Taylor/AI POD development 1990 to 2010. We have 5 PODs we are building today. It is still a low production business and we still have yet to build out a floor based on of this solution.

The Cube Office System is based on a two foot by two foot by one foot deep book shelf Module. A double wall provides a three inch space for sliding doors. With in the 2 foot module, a variety of storage unites, closet, pull out secretary unit and work stations are provided. Everything you see can be taken down, moved and reassembled using three tools with no damage to walls, floors or ceilings.
CHP Cube Office System
New York City 1998
The Cube Office system builds up from a basic 12 by 24 by 24 inch module to complete “office” scale. The cube components can contain pull down secretaries, lateral files and storage units. An “office” can be taken down and erected some place else in a few hours with out leaving a mark on the existing built architecture.
A second generation (Foundation II Series) of this system has been developed and installed in the CGEY Atlanta ASE environment.
This version of the CubeOffice, built in 2000, introduces the Gatling Post which facilitates a greater use of glass wall which can be configured in a variety of ways and angles yet easily moved. The Gattling Post is now a multi-use component of the system. This application of the system created a an enclosed KnowledgeWorker work area that can be moved anywhere in the space with relative ease. The structure is self supporting and does not depend on the existing building.
The CubeOffice System can be used to create stand alone shelving, workstations, storage systems, wall systems and PODs. It is flexible, moveable and structurally stable when constructed. It can be maintained and moved by the end-user eliminating the issues associated with scheduling and paying for professionals. The CubeOffice System is the backbone of the 1990 concept.
POD installed at Palo Alto KnOwhere Store in 1997
my work enviornmnet 1999 to 2003

To us, the Pod represents the basic amount of space and amenity that every KnowledgeWorker requires. Of course, there are several different ways to accomplish this utility and function. The Pod is one. The skin of the Pod can be fabric, wood or Plexiglas - or sections can be open. This facilitates great flexibility and adaptability.

A second generation (Foundation II Series 2000) of the Pod has been developed that allows the system to build up from one work station (a section) , roll in sections and as a single unit. The new version is made up of 50% fewer parts than the first and actually docks with our curved WorkWalls. The Pods can be arranged in clusters - or in villages as shown on the Capital Holding layout and indicated below. By using both sides of the Pod surfaces, high variety layouts can be accomplished that achieve remarkable density while at the same time creating spaces of great Prospect and Refuge and individual utility.
The Tracery elements of the Armature System act as an visual integrator, ties the villages together, provides chase-ways for utilities and one means of circulation. It also contributes to defining vertical scale. The armature is composed of moving “plug and play” components that the users can reconfigure themselves.
ArmatureSystem - 1999
Showing Foundation II components
The Cube, Pod and ArmatureSystems are components that compose the layout indicated in the 1990 sketch. The layout, itself, can be accomplished in a variety of ways and adjusted, from time to time, as required. “Smart” furniture can “know” certain rules of arrangement and “recommend” layouts based on Pattern Language values, prior successful solutions and codes. Great density and utilization can be accomplished along with a human focussed life-workstyle - all from affordable, low volume manufactured components. The technology, means and methods that facilitate this are part of our Patent and Patent Pending and other IP.
The aramture approach was further developed with the (unbuilt) Vanguard Development Center in 1995.
Vanguard Neighborhood Layout - 1995

The metaphor of streets, Alleys, neighborhoods, and so forth, creates a powerful specification in regards traffic flow and amenity. The strong architectural forms of the basic elements and armature allow a variety of different specific furniture solution-set clusters and “negative” spaces that make the niches for sitting areas, support tools, one-off workstations and small team work areas. Larger group areas are defined by Cube Office components and fold-out WorkWall systems in a variety of configurations. Prospect and refuge along with great variety and flexibility can be achieved on the horizontal plane. However, this is not enough. The constant flat, low ceiling - particularly in large spaces - is not nature to human consciousness. There is a different psychological sense in looking up and downward within an architectural space. This “sense” is important to certain mental states and, thus, to certain processes.

For years we have encourged clients to develop the vertical space of their environments. There are many reasons [link: vertical space] to do this. There are, also, several obstacles to it: codes, the limits of existing building structures, the perception that this costs more - the overwhelming paradigm of 8 to 10 feet ceilings as the “norm.” We accomplished significant vertical movement with the Orlando Management Center - 1985 (ceiling changes and floor platforms), AEDC - 1992 (ceiling changes and floor platforms) and the Palo Alto KnOwhere Store - 1997 (ceiling changes, second floor balconies and skylights). However, the most comprehensive use of vertical space was the 11,000 square foot Capital Holding project - 1992 (second level KnowledgeWorker and tecnology work-deck). Note: this was surpassed in 2007 with the Unicredit project in Turin, Itally [link: unicredit tour].
Orlando Management Center built 1985

This Center was in continuous use for 14 years. It has served three different “owners,” the Acacia Group, MG Taylor and, now, CNL. The flexibility and adaptability built-in to its layout has allowed it to accomplish three divergent missions without requiring modification. While capital intensive when first built, its durability and adaptive utility has netted a life cycle cost lower than conventional construction.


Most of the 6,000 square feet is open and adaptable - four large areas flow into one another or each can be closed individually or all at once. The three level ceiling and light coves accomplish three things: small areas are defined as part of the larger space (so that “place” is maintained with 6 or 60 in the space); the ceiling is free of the usual clutter and therefore “reads” as an active architectural element; and, the HVAC is delivered from within the coves and does not blow on people in the space.

AEDC Gossic Leadership Center built 1992. Jerry Headly, Design Development and Architect of Record. This was without question the most productive navCenter ever built in terms of organizational transformation and economic results.
This environment was built for the Air Force provided over 100 DesignShops for the Air Force, NASA, the Aerospace community, Universities, local businesses and other members of AEDC’s ValueWeb within the first 30 months of operation. The balcony above the Radian Room is the KnowledgeWorker work-loft.

Palo Alto KnOwhere Store built 1997 viewed from the second level incubation Loft. The knOwhere Stores incorporated retailing, event environments , business incubation, research and the MG Taylor corporate offices. They operated between 1996 and 2003.

Work under way at the two story high opening dome at the Palo Alto knOwhere Store center point. The entire front wall of the PA store opened up and so did the dome. This provide natural ventilation which reduced the need for heating and cooling in the temperate Northern California environment even in the center of a city.

These views clearly illustrate the power of vertical spaces. This kind of vantage point cannot be accomplished any other way. Sadly, it is too often missing in contemporary architecture. Note the technology armature over the Radiant Room. It allows for a variety of setups and the space can be set to comfortably accommodate 10 to a 100 people.

The dome cost $20,000 and was considered a great expense at the time. This will be about 100 dollars a month over the period of the lease. This simple “trick” of curved forms, low, medium and high ceilings, with the play of natural light - all interacting with appropriate dynamic rhythm - creates a space that draws people like a magnet. This area acts as the “center” of the 20,000 square foot space and turns what otherwise would have been a low, dark environment into a dynamic, exciting place to be. The dome with the first and second floor ceiling collars added less than one-half a percent to the cost of the project - netting a high architectural return on investment.

View from the knOwhere balcony under the dome to the first floor area. These spaces, on both levels, became the most popular in the environment and most spontaneous in their use. Even thought this was the most public place in the 20,000 square foot building, it was the first place people chose to do their intimate team work.

Budgets come and go - buildings last a long time. This building is forever altered and now has a value many times the investment. Spending money on “features” will not necessarily do this. The treatment has to be based on intrinsic human values and the way that people interact with space. It has to support the processes that take place within the environment.


In a world that increasingly builds bland environments lacking identity, brand and “sense of place,” people respond strongly and positively to strong architectural armatures, the use of natural light, fresh air as the weather allows (more often that supposed) and carefully selected high quality materials. This does not have to be expensive. The Palo Alto lease, including lease hold improvements and interior furnishings, was under $30 a square foot per year in a real estate market that had less than 1% vacancy at the time of remodeling, and typically rented undeveloped, unfurnished space for more than this.

The Capital Holding project also involved extensive level changes. Built on the first floor of a traditional building a two story space was created by going down five feet, opening the ceiling to the second floor concrete slab and building a loft in-between.
Capital Holding navCenter built in 1992. The view is from the 12 foot wide by 80 foot long KnowledgeWorker Loft area, looking down into several work rooms and two Radiant Rooms.
This created a variety of different spaces in what otherwise would have been a uniformly dull environment. The schema actually build a “building” within a building and allowed a progression of incremental environs as users moved in and down from the traditional areas surrounding it. While built in a corporate headquarters, and while being sensitive to the traditional detailing and materials (which were brought into the space), the navCenter “felt” like an off site experience and a world of it’s own.

The Capital Holding steel structure created an armature which acted as a conduit for electrical, LAN, multimedia and phone lines - and, supported the the KnowledgeWorkers deck from which the above photo was taken. The WorkWalls could be moved from one side to another - along with all lines - allowing the Break Out/Office space to open to another area of the Center. It was with the Orlando Management Center and this project that we laid the foundations for our technology and wiring strategies.

Throughout the 90’s, considerable technology integration was accomplished. Video, audio, computer, phone, LAN lines were developed into an adaptable “plug and play” system allowing multimedia capture, production and playback to become a real-time, integral part of the work process.
The Borgess NavCenter has Foundation 2 WorkWalls and Cubes with detachable, rolling “power poles” that “plug and play” with the wiring above, In July (1999) just a few weeks after installation, the KnowledgeWorker staff reconfigured the entire 8,000 square feet space - including A/V, power and LAN lines - in 3 hours.
The Borgess insall - May 1999

The next stage of this [link: technology integration] process involves building computer and multimedia technology into the WorkFurniture, establishing Center-to-Center RemoteCollaboration and RemotePresense, and developing the first generation of “smarts” into the system components themselves. This work is presently very high on AI’s agenda and is key to several aspects of our Patent and Patents Pending.

Most of the basic ideas that were put into product during the 90s came from my 60s, 70s, and 80s thinking. We built more, however, in this last decade than all the others before it. This allowed us a faster, more iterative, rapid prototyping process. The 2001 and 2002 installations became mature expressions of these ideas.

Even so, in the 1990s, we are just getting to the level of component design and manufacture that we envisioned when we started.

The 2002 Vanderbilt project moved further along the technology integration path by employing plug-and-play power poles, which connect to a number of bundled wire whips in the ceiling, and rolling monitor stands based on the Gatling Post system. This way a large foot print can be arranged in many different ways and equipment connected quickly without extensive rewiring. The floor is left completely free of plugs which interfere with rolling WorkWalls and furniture.

The level of component capability, that we are able to deliver today, allows a utilization of space and level of knowledge “manufacture” than can not be achieved without it. We are now able to produce an environment that utilizes Pattern Language principles, serves the full venue of individual, team and large group collaboration functions, provides greater personal space to each KnowledgeWorker than other systems, and overall, achieves greater space utilization than standard layouts using existing manufactured components.This is not an accident but the result of a decade of progressive projects that integrated design and building with operations. All this can now be done at competitive costs and in a radically reduced design-to-move-in time frame. This cannot be accomplished by design - or even manufacturing - alone. While there are issues related to the design of architecture - this is not the major problem. The furniture manufactures are capable of far more progressive designs than the present market will buy. Clients and customers want more but maintain a buying-system that makes the creation of quality workspaces almost impossible. The total system and method by which this work is done is the “problem” - I discovered this upon my entry [link: 1956] into architecture. I have devoted most of my years of work since then to discovering - and inventing - a way that consistently produces another kind of result.

AI PODcluster - ArmatureSystem - 1999

Getting to this “point of beginning” took 20 plus years since what is now MG Taylor was formed. Every step along the way was funded by a real client solving a real problem. Every piece and component went into a work environment dedicated in some way to creativity/collaboration augmentation. The design team that did this work has been small. I started the process alone and was joined by Langdon Morris in the late 70s and again in the early 90s, and then the current members: Bill Blackburn in the early 80s and who operated AI [link: ai] 1992 to 2002 and still designs for us, Gunner Kaersvang (1995-1999), Paul Lyons 1997- 2002). On the design-detailing and production end, Bryan Ross since 1987 becoming full-time manager of the AI production facility starting, 1998 to 2005. As small as AI is, it has one of the most comprehensive, and integrated, lines of work furniture solutions and design services on the market today. And, it sells more of this new kind of furniture than many of the large manufactures. As we proceed along this path, we will build an integrated design/build/deploy-use supply chain that is deeply embedded in our larger enterprise [link: mgt valueweb].

When we started MG Taylor there were no components that did what we needed - not on the utility level nor on the aesthetic level. More importantly, however, was our need to accomplish - through direct experience - the close connection between design-build-use [link: dbu model]. This is necessary for getting complex product design and intimate market fit. Because of our ability to integrate “lust-to-dust” we have been able to field solutions that multi-billion dollar organizations have failed to match. Lean production is useful, but it is Lean design/build that is necessary for true innovation. We have not been able, yet, to do this at any great scale. To do this we will have to partner with many new, large and small, ValueWeb [link: valueweb model] members that have capabilities and core-competencies that we do not. This, however, is consistent with the Network Economy concept and our own organizational strategy.

We are just beginning to see - in this newly emerging postindustrial, global economy - this kind of linkage in ValueWebs [link: valueweb architecture] and supply chains. Product improvement is both user-driven and “pushed” by a strong producer vision. Batch production is giving way to lean, mass-customization methods allowing for rapid development and evolution of products and services. The distinction between products and services is blurring. Design wins as we can see with the i-Mac, TT and Mini. Making things and delivering them to a market is the best way to design. This is employing the 4 Step Recreation Model [link: 4 step] and 3 Cat [link: 3 cat] Models as an OS for a complex ValueWeb. It involves, for us, using our product/service offerings as the factory for making them. We are intimately involved in the entire design, build use cycle.


My design work of the 90s was not just about the “what” - the product. It was about putting in place an “engine” of creation that could build the environments we wanted - which themselves, were “engines of creation.” This was, and still is, a bootstrap process. Multiple iterations to make the tool to refine the tool to make the tool. The decade of the 90s has come and gone. The work we now have in progress will accomplish many of the elements indicated in the Capital Holding sketch. [link: ch sketch] I do not see this work as the “end” - I see it as a minimum foundation for an effective future work environment. In reality, we are just getting in place many elements that should have been ubiquitous in the work environment long ago [link: reworking the workplace].


Joseki Offices, June 2002, were hand built on site by a team of students from SFIA. A mundane space was turned into a flexible environmnet expressive of the Joseki brand.

[link: future link]


When a new position is created, it acts as a meme [link: what is meme]. This “sets” the game, within which, all the variations are played out until a new game is created. The “office” - the workplace - is in the process of being recreated. Which concept of “office” gets on the “increasing returns” slope involves a critical competition. A competition of ideas, of mental bandwidth of decisional maker’s attention. It is not the money issue that matters so much as the total set of consequences that flows from the choices made and what is consequently built. How people think and work frames the range of solutions that they can see and work on. This in turn establishes the quality of what we do as a species [link: a future by design not default].


Matt Taylor
Palo Alto
February 20, 1999


SolutionBox voice of this document:

click on graphic for explanation of SolutionBox

posted: February 21, 1999

revised: October 15, 2010

• 20000603.905342.mt • 20001101.768821.mt •
• 20010204.569812.mt • 20020612.640091.mt •
• 20020617.666622.mt • 20020619.123400.mt •

• 20101015.211209.mt •

(note: this document is about 98% finished)

Aspects of the system and method described are Patented and Patent Pending.

Copyright© Matt Taylor 1999, 2000, 2001, 2010 , 2002

Search For:
Match:  Any word All words Exact phrase
Sound-alike matching
From: ,
To: ,
Show:   results   summaries
Sort by: