“The Science of Digital Fabrication” at MIT’s Center for Bits and Atoms
Alvise Simondetti visited his alma mater, MIT, last week to attend a conference on “The Science of Digital Fabrication” at the Center for Bits and Atoms, organized in partnership with the Executive Office of the President of the United States. The event was an opportunity to bring together leaders from research and governmental sectors to discuss the future of digital fabrication from the perspective of a variety of scales, methods, and applications.
Digital fabrication technology originally arose out of MIT in the late 1940s with their development of the first computer numerical control (CNC) milling machine. This invention initiated a growing variety of subtractive manufacturing processes, or machines that would take a large piece of raw material, and methodically remove areas with tools such as lasers, waterjets, and wires, to achieve the final design. Additive manufacturing, or the deposition of material to form the finished product through processes such as 3D printing, has been gaining attention recently also due to the potential to produce less waste than reductive processes.
The workshop took place on March 7th, and was divided into three briefing areas: Materials and Mechanisms, Processes and Workflows, and Policy and Programs. Attendees then broke into working groups to further discuss and develop specific areas of interest. The event concluded with the 7th annual Goldstein Lecture, given this year by Swiss architect Matthias Kohler.
Materials and Mechanisms
Neil Gerschenfeld, head of MIT’s Center for Bits and Atoms, opened the Materials and Mechanisms briefing session. Neil’s research focuses on the relationship between digital and analog processes, and the emerging feasibility of “personal fabrication,” providing non-technical users with the skills and technology to produce objects to suit their specific needs. He began with a provocative insight, that 3D printing will be the “microwave of the future kitchen,” in other words, that it will be integral to the future of the field, yet just one small piece of the transition from analog to digital fabrication. Current digital fabrication technologies include subtractive methods such as CNC routing, laser, and waterjet, as well as additive techniques such as extrusion deposition, granular materials binding (sintering, electron beam melting, and plaster-based and plastic-based 3D printing), lamination, and light polymerization. Jo Newton, founder of TechShop, provided the additional insight that digital fabrication will follow a similar trajectory as desktop publishing: beginning as a service offering of specialized shops, and ultimately transitioning into a ubiquitous element of everyday experience.
Materials and Mechanisms session
The session focused on digital fabrication at the nano (DNA), micro, meso, macro, and mega (houses) scales. It was insightful to have a collection of cross-disciplinary and cross-scale industry leaders gathered together to share their experiences, as many “lessons learned,” including tools, processes, and metaphors, could be applied across the different scales.
A segment of the conversation focused around digital fabrication being the “next industrial revolution,” which will increasingly involve the production of dynamic, interactive objects, or programmable material with properties that adapt to accommodate specific needs. The possibilities of digital manufacturing allows for a departure from the production of systems of thousands of static components (ex., an airplane wing), and a movement toward the design of the materials themselves with specific intrinsic properties. For example, digitally pre-organizing a material to contain varying degrees of porosity in different areas can modify the material’s buckling properties. This will enable the holistic design of objects more similar to the wing of a bird, combining information and responsiveness with the materiality of an object. This revolution in total, rather than elemental, fabrication will enable electronic transistors or RFID tags, for example, to be manufactured directly into materials, resulting in the production of truly smart objects.
A critical insight from this session is the need to balance complexity with efficient design processes. For example, the design of jet fighters is increasing exponentially in complexity. Whereas the last generation of jet fighters took two years to design, and the current generation takes about seven years, it is projected that the next generation will take 20 years to fully design, due to the need to evaluate adaptable/programmable materials within a range of possible scenarios. In this example, digital fabrication is creating, rather than solving a design challenge, as technology will have leapfrogged these jets by a wide margin, rendering them obsolete in advance of their completion. The solution can be found in the mapping of the human genome: open innovation is necessary to reconcile complexity with efficient design.
MIT digital fabrication workshop
Another insight from the Materials and Mechanisms session was that, for the first time in history, design is moving into spaces where we lack the ability to intuitively solve the problem at hand. While increased automation is allowing a drastic increase in the complexity and sensitivity of our designs, we are becoming increasingly reliant on technological solutions to solve the problems that the same technology is creating.
Other speakers in this session included Saul Griffith, inventor, who presented on the history of the technology, Joe Jacobson, head of the MIT Media Lab’s Molecular Machines research group, who spoke on fabricational complexity, Ned Seeman, inventor of the field of DNA nanotechnology, who described his research on self-assembly of nanoelectronics, and Skylar Tibitts, director of MIT’s Self-Assembly Lab, who gave examples of his self-assembling objects using magnets and applied kinetic energy. His self-assembly research is of particular relevance to Arup because of its application at the mega-scale of the built environment.
To conclude the session, Larry Sass, professor in the Department of Architecture at MIT, made the critical point that in order to enable a more fluid design and fabrication process, the same set of “instructions” must work both at the scaled level for prototyping and ultimately at the full scale for final construction. When design specifications are scalable from model to completion, the act of prototyping becomes an integral tool in the design process itself.
Processes and Workflows
This session provided techniques for increasing the fluidity and efficiency of design processes and workflows. This fluidity can be achieved by using automated extraction of design parameters from object functionality, as well as novel scanning techniques, which directly capture the information required in the manufacturing process, increasingly sophisticated functional representations that that enable direct fabrication, and a departure from methods of data communication that inhibit speed and complexity.
The Processes and Workflows briefing session provided some interesting insights on the state-of-the-art of 3D scanning, printing, and folding, as well as the optimization of various fabrication and motion control techniques. Wojciech Matusik, professor at the Computer Science and Artificial Intelligence Laboratory at MIT, spoke about his research on simulation and optimization. He gave the example of computing kinematic movement for animated movie characters, and detailed the process for fabricating physical characters using an automated design process to characters with a similar range of articulated movements.
Philip Withers, professor of materials science at the University of Manchester, explained his research on applying computational tomography (CT) scanning outside of the medical realm. Traditionally, 3D scanning has been limited to surface scanning, but CT scanning provides the added benefit of volumetric assessment, or “seeing through” an object. He has applied this methodology to fossils embedded in rocks, enabling an accurate 3D reconstruction of the skeleton of dinosaurs.
Philip Withers speaking on volume scanning
Matthew Keeter, grad student at the Center for Bits and Atoms, opened a discussion on design representations and interfaces by examining the weaknesses of CAD/CAM workflows. He explained that many of our current design tools rely on boundary representation (B-REP), which identifies objects based on the limit between surfaces, but has difficulty producing smooth fabrication output. Functional representation (F-REP), or the description of an object based on a single function, has the ability to produce more accurate 3D objects, yet it was abandoned in the 1970s due to the inability of then-current computing technology to manage the complex screen representation and renderings. Computers are now able to handle F-REP, enabling direct fabrication of designs. Software such as Rhino are keeping up to speed with this trend in representation with the development of F-REP plug-ins.
Additional presentations included Sanjay Sarma, who spoke about a sophisticated method of path planning which computes an optimized “line of sight” for an undercut to create efficiencies in fabrication. This is another example of research that was abandoned in the 1990s due to limitations of technology, and is now providing opportunities for the future. Nadya Peek spoke on motion control and the development of a global network of FabLabs. Informed by her global experience with FabLabs, she challenged the usage of methods of communication with machines which are constrained by antiquated simplistic g-code, and stressed the necessity for the industry to embark into more sophisticated communication of data. Jennifer Lewis spoke on high-precision multi-nozzle array printing, which can print using multiple materials simultaneously. She noted that this allows users to print “function as well as form,” which can be illustrated through applications such as the fabrication of materials with embedded computing technology. The current array capacity is 8×8 nozzles, or 64-bit, with 128-bit under development and 256-bit projected for the near future. Multi-nozzle array printing will likely undergo an exponential increase in speed and material complexity in the coming years.
Policy and Programs
The third session featured representatives from different governmental agencies, with the intent of sparking synergies and maximizing the potential for federal involvement in emerging research and technology. John Slotwinski, delegate from the National Institute of Standards and Technology (NIST), provided a compelling argument that the standards with which we choose to measure new technology play a role in the development trajectory. For example, if we measure digital fabrication technology by speed, we may limit the growth of other aspects, such as precision or complexity. He also spoke on material choices, noting that NIST has decided to prioritize metal as opposed to polymers as a promising material for digital fabrication.
Paul Eremenko, from the Defense Advanced Research Projects Agency (DARPA), challenged the design/build/test/redesign workflow that designers and engineers currently use. He explained that design complexity is projected to increase sevenfold in the coming years, and that we will no longer be able to afford this extended design process. A higher level of abstraction is required for future design processes. Similarly, the example of BBC’s innovative usage of semantic technologies (use of metadata and automated tags) to arrange news stories automatically across multiple websites, which drastically reduced the number of reporters necessary to fully cover the Olympic games. As complexity rapidly increases, such design process optimization will be crucial for the integrated workflow of multidisciplinary project teams at Arup.
Paul Eremenko speaking on design complexity
A critical insight from LaNetra Tate (NASA) is that seven out of nine of their research centers are currently engaged in research relating to 3D printing. Digital fabrication techniques hold dramatic opportunities for construction processes in space: because certain 3D printers are capable of using dirt, sand, and other materials from the surrounding environment, only the printer itself and the bonding agent would need to be transported to enable large-scale construction trials on the moon, as early as 2014.
Other speakers included Kelly Visconti, (DOE), who posed the question of how much energy is embedded in manufacturing processes. She noted that wind turbines, made of thousands of components, have difficulties recuperating their embedded energy, but that digital fabrication brings with it the opportunity to positively impact the waste and energy intensity of manufacturing. A representative from the Department of Homeland Security (DHS) discussed geoprinting of infrastructure, including the benefits in emergency situations, such as hurricanes and other natural disasters. DHS was also aware of the potential threats of digital fabrication, such as the possibility of printing components for illegal weapons, or the modification of DNA for harmful purposes.
The final three speakers were introduced as the “sausage makers,” or those who were getting their hands dirty on the front lines, making change happen. Dale Dougherty (Make) posed the critical question of what the tipping point will be to enable widespread adoption of the technology. Ensuing discussion proposed that the tipping point will be the establishment of 3D printing as a service, coupled with an intuitive interface that is able to auto-check designs for critical requirements such as stability, rigidity, or watertightness. Vincente Guallart (chief architect of Barcelona) spoke of the importance of bringing manufacturing back to the local level, and the possibilities associated with open innovation. Together with the mayor or Barcelona, he championed the first FabLab to be implemented outside of MIT several years ago, and proposed that a distribution of digital fabrication centers around the city, each with a specific city-related task, will hold great opportunities for empowering people to devise their own solutions to municipal needs. Congressman Bill Foster concluded the proceedings of the day by thanking participants for gathering to discuss the future impacts of digital fabrication. In 2010, Bill proposed the National Fabrication Network Act, a revolutionary idea to empower all Americans with the opportunities created by access to digital fabrication tools.
Vincente Guallart speaking on digital fabrication in Barcelona
Goldstein Architecture, Engineering, and Science Lecture
The concluding lecture, “The Design of Robotic Fabricated Architecture,” was given by Matthias Kohler. He spoke about the expanded usage of robots in architecture beyond traditional applications such as increasing safety, efficiency, and economy. The global accuracy that automation brings to the construction process is well beyond feasible human replication, and could even absorb much of the human error along the way. One example that he outlined was the laying of bricks offset at a one-degree angle to produce a complex, textured surface.
How might digital fabrication impact your life?
Let us know: twitter.com/arupforesight
For the full program of events and speakers, please visit: http://cba.mit.edu/events/13.03.scifab/index.html
Written and researched by Katherine Prater