Fostering Innovation, MIT Media Lab Style
Ryan C.C. Chin

Imagine what would happen if you locked up a small group of architects, mechanical engineers, computer scientists, visual artists, urban planners, and an MD and asked them to work together? Chaos of course! Such differing research and working styles could hardly be compatible unless they were given a compelling problem to solve. Throw in the fact that these are some of the brightest and most talented students at MIT, and the situation becomes even more complex. The MIT Media Lab's concept car project with General Motors and the distinguished architect Frank O. Gehry is just such an example. In constructing a vivid picture of how innovation may be introduced to any enterprise through multidisciplinary collaboration let's look at this Media Lab project and three categories of innovation: (1) research methods, (2) design thinking, and (3) making.

Innovation in Research Methods
Innovation can come from anywhere. Often an outsider can shed light to a problem, thus more is more when it comes to ideas. Therefore, it is much better to be inclusive versus exclusive. The ability of the group to work together is also important. The differences should create a comfortable tension, not deadlock. The key is setting up a framework for innovation to happen. In the concept car project, we did not begin with a problem but with a premise: "reinvent the car." This open-ended approach allows for a wide range of interpretations which enables each team member to participate and to take ownership of ideas.

Re-examining existing problems is a simple way to initiate innovation. We looked at current cars and our sub-groups focused on chassis, interior, vehicle control, envelope, and materials. Collectively they would gather research and soon become advocates for their topic areas.

In the case of interiors, the group began systematically mapping out every possible seating configuration for a 4 passenger vehicle. Many of these configurations are well-known and currently in use. However, several compelling and even ridiculous configurations materialized, such as the 1+2+1 configuration or the 1+1+2 configuration. The 1+2+1 configuration we affectionately called the "Rock Band" configuration. “Rock band” seating configuration: Will Lark, MIT Media Lab.

Asymmetric configurations were heavily scrutinized because of aerodynamics and packaging concerns.

Many non-standard configurations have been proposed in the past, but have not materialized because either technology or the culture was not ready. The "Rock band" seating configuration for many assumes that the driver would be in the front and therefore would not be sitting next to the front passenger.

Innovation spurs more innovation.

Design is all about making decisions. Making a visual map of possible seating configurations is useful only if a metric is established to determine which ideas are useful and which ones are not. But dismissing a potentially fruitful idea too early can halt the creative process. A concern that many designers had with the Rock Band configuration is that you could not accommodate routine social norms. However, a couple can sit together in the Rock band configuration if a drive-by-wire vehicle control system was available. The front seat can fold down and become storage when there are only two passengers in the vehicle. By-wire technologies have recently become practical in the automotive space. By-wire has become a design parameter in this project.

Innovation in Design Teaching
Design and engineering synthesis is a problem at all levels of academia and industry for reasons of communication, interfaces between disciplines, and the sheer size of some organizations. Creating teams that tackle specific issues is a well-documented process, however, what happens when the issues migrate into large realms of expertise? Complex and mature products like automobiles require the collective knowledge and insights of hundreds of individuals in diverse areas of expertise.

We took this opportunity and created a two-step process. First, we developed research groups that became advocates in their areas. Then we disassembled those teams and created new ones as mixtures of these advocates, being careful not to combine too many of the same disciplines in one group. The challenge for these teams was to carefully weigh the issues they believed in and to also suggest solutions. Often there was a clear division of opinion. Our answer was to simply follow two or more solution paths until it became obvious which direction was better.

Often at MIT we do lock people up in rooms, give them some food, and tell them not to come out until they have come up with something. Unfortunately, we don't have a chimney with white smoke as they do in the Sistine chapel during conclave. As opposed to the papacy where the solution is a singular selection, our results are often a multitude of answers.

Our envelope group was in charge of the vehicle exterior. They postulated that painted sheet metal was a problem for pedestrians. Of course, sheet metal has its adherent advantages of strength given its weight, however, for pedestrian impact; it is not a forgiving material. One possible solution is to cover the whole exterior with inflatable airbags! What would that be like? Aside from the aesthetic departure of conventional stamped sheet metal, a number of advantages emerge such as (1) impact mitigation, (2) design freedom, and (3) a new set of material properties such as scratch resistance.

New safety laws have been passed in Europe recently regarding the interface of the vehicle exterior and pedestrians at the point of impact. These laws will eventually migrate to North America and have unexpected design consequences. The outer geometry of the exterior would be constrained to specific angles thus restricting designs to a particular form language.

Our innovation was to solve the problem through a material choice rather than a mechanical solution. Innovation in design teaching allows for such explorations because the design path is never straight. Often we discover new opportunities as a result of suspending a few parameters at the expense of investigating a new area.

The greatest single deterrent to innovation is the fear of being shot down. Cultivating an environment that rewards “doing” is key. Often ideas are faced with a firing squad jury and do not even have the chance to be fully formulated. Building upon the ideas of others is encouraged as it also allows innovation to flourish. This fosters a sense of camaraderie and a trust in the ideas of others.

Making decisions is better than not making them because otherwise the process is paralyzed. If the decision was not right, you have at least explored the options and that will inform other choices to be made. Innovation through design teaching is the rigor placed in setting the problem which facilitates design decisions and tangible solutions. To simply debate is not good enough, but to take a leap of faith, be it a sketch, a physical model or a list of characteristics, is necessary.

Innovation through Making
In 1998, Professor Neil Gershenfeld established a course called “How to make almost anything” at the Media Lab. This intensive course was an introduction to all the fabrication techniques available in our facility. Students and faculty could learn how to program a chip, use a laser cutter, mill a piece of aluminum, make a 3D model, use a 3D printer, and much more. Their new skills would be triggered by specific design themes. For example, graduate student Kelly Dobson created “Screambody.” The Screambody allows the user to record a scream for later release. During the course, Kelly learned to fabricate all the electronic components, cut the fabrics, programmed the microchips and created the device. Innovation was enabled by making. http://web.media.mit.edu/~monster/screambody/ (1998).

A distinct shift in focus at the Media Lab and in industry has occurred in the last 10 years. Instead of focusing on abstracting electronic content from physical representations, the Lab is now experimenting on how electronic information overlaps with the physical world. Tackling the physical world has been enabled by the Lab's collection of computer-controlled fabrication tools. The process of making is extremely diverse. Early sketches are best done with pens and pencils, physical models can be made out of basic materials like clay or cardboard, gears should be cut precisely by using CAD/CAM tools like a CNC milling machine.

One such program that we have used to expedite innovation is CATIA V5. It is the same program that Frank O. Gehry uses to create his complex and envelope-pushing designs like the Disney Concert Hall in Los Angeles and the Guggenheim Museum in Bilbao. CATIA has enabled us to make and to also test our ideas, either virtually or physically. Our use of CATIA can be roughly divided into parametric modeling and master assemblies comprised of parts.

The value in parametric design modeling is taking a number of known constraints and producing a number of possible solutions. We call this the “solution space.” The seating configurations are an example of a parametric solution space. Take 4 seats and constrain that with a given interior footprint. This yields dozens of possibilities. This is the first step in “making” because from this model we can visualize the designs by creating a rendering or animation. We can also take this model and send it to a rapid prototyping machine like a 3D printer and make physical models. Building smart parametric models also reduces the number of iterations ultimately needed. Non-parametric modeling programs require you to remake a model even if a small change occurs.

Countless models, drawings, sketches have been made in the two years since this exploration began. Hundreds of designs have been generated, critiqued, and developed. Every idea or thought has been of either immediate or latent value because they have added to the collective intellectual capital of the group. In the very beginning of this project, as we were discovering our way, the brainstorming meetings were quite contentious. By the end of the process, we stopped talking and started making. If pictures are worth a thousand words, then a 3D model should be worth ten thousand, and a rendered animation of that model should be worth a million.

Links:
http://www.media.mit.edu
http://cities.media.mit.edu

About the Author:
Ryan Chin is PhD Candidate in the Smart Cities group at the MIT Media Lab. He holds undergraduate degrees in Architecture and Civil Engineering from the Catholic University of America, and a Master of Architecture, and a Master of Science in Media Arts and Sciences from MIT. He coordinates the design research for the MIT Media Lab concept car project with General Motors and Frank O. Gehry.