Q-BA-MAZE

Minnesota Design Mafia

Seven years ago, this red-headed fellow set in motion a series of events that led to the creation of Q-BA-MAZE:

He is Thomas Fisher, the Dean of the College of Design at the University of Minnesota (CDes UMN), pictured here at the opening party for Here By Design III watching a ball wiggle its way through the cubes.

In early 2001, he organized a gathering for UMN architecture alumni living in New York City -- a social and networking event attended by about 15 people. I was living in Brooklyn and had just started my architectural rendering business. The gathering was at the Manhattan office of EEK Architects, where UMN alum Peter Cavaluzzi is a principal. I offered Peter my rendering services and came back to his office a couple of weeks later to show him my portfolio. It took a while to build the relationship, but because of the implicit trust from the Minnesota connection and the whole Mid-Western work ethic thing, I eventually had fairly regular rendering assignments from Peter. The intermittent nature of the work was exactly what I wanted and it was during one of the lulls between assignments that the inspiration for Q-BA-MAZE hit.

I moved back to Minneapolis in 2003 and continued to get rendering assignments from Peter at a distance. Then in summer 2004, EEK was competing to become the master planner for Project City Center -- a $4 billion casino project for MGM on the strip in Las Vegas. I did the initial renderings and the project just kept expanding to include more renderings and eventually videos. EEK won the commission and the renderings ended up in the Wall Street Journal and the videos played on the CBS Evening News. When my part of the project was complete and my final paycheck came in I calculated that I had enough money to put everything else aside and begin full-time work on Q-BA-MAZE.

Here is the rendering that was published in the Wall Street Journal (in this case, computer-generated buildings composited into an aerial photo taken from a helicopter):

For six months I simultaneously researched and wrote the business plan, designed the cubes, initiated the patenting process and incorporated the business. All of the rapid prototyped cubes on display at Here By Design III were made during that period. At the end of this process I had a product and a business concept that was sufficient to convince investors to come on board so that we could take the expensive next steps of bringing Q-BA-MAZE to production. I've always thought it funny that I got all of the pre-investor money for this risky bet from Las Vegas.

Above is another rendering of the project that is 100% computer-generated (all of the tiny specks at the bottom of the image are people -- which is to say, this project is huge).

There was a time when I thought design was about things, but I am learning more and more that it is about people and relationships. Tom and Peter are just two of hundreds of people who in one way or another have made Q-BA-MAZE possible. And so:

This is post #6 in a series on "The Making of Q-BA-MAZE"

Backstage: Mini Photo Studio

This post is a little look backstage at Q-BA-MAZE to see where the photos come from. This first photo is mostly complete. I'd still like to see how it looks with the background completely whited out. But I'll have to find time for that tedious Photoshop exercise another day. If you want to build this configuration, here is the Excel plan.

Just back up the camera a little bit and a bunch of equipment sneaks into the frame. Here I am working on structures for an upcoming cantilevers and helixes post I'm planning. I want to get some extreme shots that show how structures can tip and spin without coming apart.

I'm using this new compact fluorescent lighting system from Interfit. The light is very close to natural light and the bulbs maintain their color for thousands of hours, so I should have color consistency over time. These lights run SO much cooler than other types of lights which is a real bonus on a hot summer day. I'm shooting with a Canon Rebel XTi using the standard zoom lens that came with the camera. Note the Manfroto tripod and Manfroto 329 RC4 quick release head. It is very handy to be able to easily disengage the camera from the tripod and later reattach it and still have the same shot composed. I also still use my Nikon Coolpix 995 from the digital stone age because it still does what I need it to do and I have this insane 72mm wide angle lens for it that helped me get all of the backstage equipment into this last photo.

This is the basic setup. There is a bunch of experimentation with the lighting to follow. The filtered lights are for making the background bright and white while the bare bulbs are for creating a bit of sparkle in the cubes.

My next step on this shot is figuring out how to best shoot a stop action film of a cantilevered, zig-zagging, helixing structure. And now that I"ve made this backstage view, I"m thinking I like the rough do-it-yourself look complete with the orange Pony clamps. Four of these clamps are holding a long sheet of paper in place. The paper begins on the table top and curves up behind in order to create the seamless white background.

I've lost any sense of logical order here, but I consider the photography a part of the "making of" series so:

This is post #5 in a series on "The Making of Q-BA-MAZE"

SKETCHING: The ongoing importance of free-hand drawing

As irreplaceable as 3D CAD is to the design process, sketching by hand is still a necessary skill. Architects have a phrase, "don't talk about it, draw it." Words will gloss over all the difficulties, a drawing exposes the difficulties and forces a resolution.

The sketch included here is a combination of the "just draw it" philosophy and a method I learned in a design course from architect Tom Meyer of MS&R Architects while I was a student in the University of Minnesota architecture school. He calls this "talk drawing." The idea is to record design thinking in both drawings and words side by side on the same piece of paper -- and to actually talk outloud. If you leave and come back to this "talk drawing", you have a rich source of material to review to get your mind back into all of the myriad issues that can be at play. The visual-spatial and the verbal-analytical parts of the brain are both fully engaged in this kind of drawing.

This particular drawing deals with a number of issues that came up in the middle of the rapid prototyping process (during the decision to decrease the size from a 2" to a 1 1/2" cube). Among the issues are the supporting ribs inside the cube, the spherical shape of the interior, and the unification of the side joinery elements into a single horseshoe shape.

At the top of this post, I include the detail zoom into the drawing mostly because pencil on paper looks so amazing. You can SEE the thought process in the lines, the erasures, the pressure on the pencil, in a way that is just never revealed in CAD drawings. 

If you look closely, you'll see that this was drawn on the back side of a Private Placement Memorandum I was writing in parallel with the design process, in order to raise money for the business. I was at my local Dunn Bros coffee shop reviewing the PPM and did not have a computer at hand. An ability to draw free-hand allowed me to get some important thinking done when inspiration hit. But another reason for approaching a design problem with sketching and plaster and CAD and rapid-prototyping, is that different things are noticed, different insights gained in each method. So it is not so much that one process is more important or better than another, I think the important thing is to have a multi-faceted process that will help bring out all of the issues.

This is post #4 in a series on "The Making of Q-BA-MAZE"

2D/3D CAD DRAWING: Coordinated Precision

I do all of my CAD (computer-aided design) drawing in 3D with software called Microstation (Microstation is not typically used for product design, but it is the software I know from my years in architecture, so I have stuck with it). I do not need to draw 2D drawings, instead, the software automatically generates the many 2D drawings from a single 3D design. This saves a lot of time. It also means that every aspect of every drawing is perfectly coordinated, because it all refers back to the same single 3D object.

The drawing on the left is actually part of the submission to the US patent office. The software assisted me in making a very thorough submission of many variations of the Q-BA-MAZE concept.

The same 3D file that produced these drawings also produced the rapid prototypes. And when I was getting initial bids from injection molders in Canada, I was able to tell them the exact volume of the parts, again, by analyzing the same 3D design file.

The surface number 600 in fig 2B is the concave-up sphere, on which the cascading balls will pause while rocking back-and-forth. This sphere is placed in the same location in every type of Q-BA-MAZE cube. Only the exit configurations out of the cubes vary. The precise alignment of the top of the side joinery and the intersecton of the cube wall can be seen in fig. 2I. But without getting into every detail, I will summarize by saying that these drawings show that the cubes contain an abundance of detail and every one of those details impacts the next because there is almost no tolerance in these parts. The only option is that they fit perfectly. CAD software made this possible.

This is post #3 in a series on "The Making of Q-BA-MAZE"

RAPID PROTOTYPING: Design Gets a Reality Check

Rapid-prototyping is a means of "printing" working 3D objects directly from a 3D computer file.

Q-BA-MAZE has an underlying geometric simplicity since it is just a system of cubes. But an endless complexity lies in the details of these cubes -- the internal geometry that slows the cascading balls and the  joinery that allows daring cantilevers required months of design investigation.

Rapid-prototyping was at the heart of these investigations. I made at least eight generations of the design in this way. With each iteration, lessons were learned and the design improved for the next round. I usually made about 6 cubes in each generation, enough to test the joinery and the movement of the balls through a simple configuration.

This gray cube is the first generation. The side joinery is a kind of split dove-tail connection that reveals my background in woodworking. This joinery was not rigid enough. The bottom edge of the cube had an unacceptable "accordian" action.

This red cube is the second generation. The side-joint now has a "hook" shape. The "accordian" on the bottom edge is gone. The interior of the cube is a cylinder in an attempt to make a shape that would keep the balls rolling. The cylinder and the uninterrupted bottom edge also work together to increase the rigidity of the part.

This is the "solid" version. I had noticed that Philippe Starck's Ghost Chair had polycarbonate resin in excess of 1/4". I thought making the walls so thick would be the way to give the design quality through solidity and a beauty similar to crystal glassware. The thickness also allowed the corners of the cube to become rounded and pleasing to the touch. This was a bomb-proof design that proved to be far to expensive to manufacture because of the sheer volume of resin required and the size of the press to run the mold. It also turned out to just seem heavy and clunky rather than luxurious.

In this generation the scale of the cube is reduced from a 2" cube to a 1 1/2" cube in order to make the cubes fit more comfortably in the hand. And four internal ribs are added to give the part strength and solidity through engineering rather than heft.

  In this final rapid prototype generation (I've skipped over a couple of generations in which the developments are very subtle), a concave-up sphere now forms the interior geometry of the cube from edge to edge and the side-joint is unified as a "horseshoe" rather than being split in two parts. The round corners of the "solid" version have returned to the design to make the cubes pleasing to the touch, to increase surface area with the interconnecting bottom-pins, and to ease the flow of the hot liquid resin during molding. The four ribs have also been extended upward, in the upper half of the cube, to form "buttresses" or a kind of box-beam in every corner. Satisfied that the design now had the right form, feel, and function, the design was ready for making the production tools for injection molding.

Here is the final molded polycarbonate part -- embodying the lessons of the many generations of rapid-prototypes that came before it.

This is post #2 in a series on "The Making of Q-BA-MAZE"

INSPIRATION: Grandpa's Hand-made Marble Run

With several years of effort to make Q-BA-MAZE a reality, I have a long answer to the question, "How did you come up with Q-BA-MAZE?" But that is the 99% perspiration part. The 1% inspiration is pretty simple and the subject of this post:

Living in New York City, I jumped out of an architecture career to start a business making computer generated images of buildings for architects. Crazy. But somehow it worked. And as a bonus (not a Wall St. bonus, but a bonus nonetheless) I had some time and the tools for my own design work.

One day, sitting in my apartment/office I remembered this marble run my grandfather had made many years ago. I loved playing with this as a boy on visits to Grandpa's house. I also loved Lego and building elaborate structures with wooden blocks to make tunnels for my toy cars. All of these favorite toys of my boyhood were in my mind and I wondered, "What if the hand-carved blocks in Grandpa's marble run were loose and could be continually reconfigured into different pathways?"

So I fired up the computer and drew two rectangular blocks, each with a spiral groove, one right-handed and the other left-handed. I made a whole set of these blocks and played with them in the computer, making the structure shown here. But the pathway just wouldn't go where I wanted it to. There was something inherently limiting about the design. I wanted the ball to be able to twist and turn anywhere, but the rectangular shape and the single entrance and exit in each block did not provide sufficient design freedom. Then I realized that a CUBE with more entrances and exits would both simplify the system and increase the design freedom for the user.

To test the strength of this cube concept in reality, I made a plaster prototype and ran marbles through it. The double-exit blocks really clinched it. They created such anticipation about where the marbles would go. I wasn't sure about how to proceed from there, but I knew I wanted to pursue this idea.

This is post #1 in a series on "The Making of Q-BA-MAZE"