Description

The overarching vision for the Center for Rapid Automated Fabrication Technologies (CRAFT) is to develop the science and engineering needed for rapid automated fabrication of objects of various size up to mega-scale structures such as, boats, industrial objects, public art and whole building structures. To achieve this, CRAFT develops a unique academic environment blending fundamental research with the development of large-scale engineered systems, incorporating environmental, regulatory, labor and economic expertise; partnerships with materials, equipment, construction, architecture, real estate, software and manufacturing industries; and interdisciplinary graduate and undergraduate education.

The grand challenge for CRAFT is building a custom-designed house in a day while drastically reducing the costs, injuries, waste and environmental impact associated with traditional construction techniques. The vision is a revolution in housing construction, whether it be to provide affordable housing for the 30 million U.S. households facing cost burdens or overcrowding; emergency housing for victims of disasters; extraterrestrial buildings constructed from in situ materials; new styles of housing based on curved organic designs rather than straight surfaces; or inexpensive first ownership housing for an emerging middle class in the developing world. With national construction-related expenditures currently totaling close to $1 trillion annually, the potential impact is enormous.

Over the past 5 decades, there has been little change in the basic materials and methods used in construction. Wood products are still used for residential construction, while steel and concrete are used for taller buildings. Also, the highly manual basic methods of placement and assembly have been essentially constant. Improvements have occurred largely in construction equipment and in secondary components such as windows and exterior siding. Innovations such as aerated concrete and fiber-reinforced and polymer-modified concrete have extended the use of traditional materials, but remain outside the realm of standard practice.

The industry is highly fragmented and is dominated by small-to-mid-sized companies. This situation has stifled advances and innovation in new materials and production practices. There is little incentive for introduction or development of new materials or construction techniques. Companies use the same materials and techniques as the competition to survive due to the lack of an economically viable alternative, and building code compliance allows little room for experimentation or innovation.

Ordinarily in the industrialized countries manufacturing automation for products such as shoes, clothing, cars, etc. results in production cost that is roughly 25% of manual methods. Efficient automation of construction should result in similar cost reduction. We expect this reduction to be realized through a) rapid time-to-market, which drastically reduces construction financing costs, b) savings through near elimination of material waste, and c) major reduction in labor related expenses. Furthermore, while most other products may be imported or outsourced, construction is an industry that must remain indigenous. It is therefore imperative that we devote a concentrated effort to automation of construction.

The CRAFT team has made one of the most significant recent advances in the construction industry in the past 50 years. The Contour Crafting process, invented by Prof. Behrokh Khoshnevis, affords a versatile platform for a major paradigm shift in mega scale fabrication and the introduction of new, high-performance construction materials. The automated process can be used to build intricate, innovative structures rapidly and with minimal extra material cost, a critical factor when attempting to introduce new materials. While virtually every other sector of industry has experienced dramatic improvements in efficiency through automated manufacturing and technological innovation, the construction industry has not. The industry needs new materials combined with more efficient processes to reduce labor and fabrication costs.

Contour Crafting (CC) is a hybrid fabrication method that combines extrusion for forming object surfaces and filling to build the object core in layered fashion. As shown in the upper left corner of Figure 1 , the extrusion nozzle has an adjacent trowel. As the material is extruded, the traversal of the trowels creates smooth (2 micron has been achieved) outer surfaces on each layer. The nozzle can be deflected to create non-orthogonal surfaces such as domes and vaults. Co-extrusion of multiple materials is also possible. For example, plaster as the outer surface material and concrete as the core structural material may be co-extruded by the CC nozzle.

Early and current developments, funded by three NSF grants, focus on a small scale (middle and lower left parts of Figure 1 ) but reveal the potential of CC in construction through experiments with (among other materials) clay, plaster and concrete. And we are currently creating and testing full-scale concrete wall sections (top and lower right sections of Figure 1 ).

Contour Crafting falls within the class of Rapid Prototyping, or Solid Freeform Fabrication, technologies, but the use of thick layers plus troweling to provide a high quality surface finish and rapid fabrication time affords a scaleup pathway to the rapid fabrication of large building-sized structures that no other technique provides.

Conventional methods of manufacturing automation do not lend themselves to construction of large structures with internal features. This explains why the evolution of construction automation has been slow. Automation has been used to a certain extent in pre-fab housing, which has not received wide acceptance due to low design flexibility and high transportation and assembly costs. There has not been much attempt for on-site construction automation other than in Japan where generally non-integrated robots perform specialized tasks that mostly aim only at mechanizing the conventional manual approaches (e.g., material handling robots, welding robots joining steel columns and beams, robots for spreading and smoothing concrete, brick-laying robots, etc.). As a result, successful integration has been difficult and expensive. Development of entirely new automated fabrication paradigms, such as CC, could mitigate many of the problems that the industry is facing.

Tapping the full potential of this technology requires an interdisciplinary center. We will need to develop large-scale engineering systems consisting of materials delivery and extrusion systems, multiple heterogeneous robots working together, and software that can take fabrication automatically from design through construction and inspection. A broad range of interdisciplinary research is required spanning multiple engineering disciplines, computer science, materials science, architecture and the social sciences on these hardware and software components, on the basic materials and structures that will be involved, and on the societal impacts of the revolution in construction that these systems will enable. However, it all must fit together into systems capable of: (1) rapidly and automatically fabricating mega-scale structures such as single-family homes, and (2) laying the groundwork for the transition of this technology into actual use.

CRAFT Research Strategy

Figure 2 presents the overall research strategy for CRAFT, partitioned across three planes. The Technology Demonstrations plane is home to the application testbeds. CRAFT’s engineered systems focus is on mega-scale layered fabrication. Within this focus there is a grand challenge building a house in a day that will serve as the primary testbed for the center, its engineered systems and its research. In particular, developing systems capable of succeeding with this grand challenge requires integration across the (sub)systems that will be developed by the three technology thrusts and the full spectrum of fundamental research. Additional testbeds in areas such as industrial parts and molds, public art, and extraterrestrial construction will be considered as additional industrial or other governmental funding becomes available for them.

The Technology Thrusts plane is home to three thrusts that will both produce the (sub)systems required by the grand challenge and organize and define the barriers to be addressed by the fundamental research. The Extrudable Materials and Fabrication (EMF) thrust will research and develop materials, extrusion systems, and structures built by extrusion of materials. The Modular Components and Assembly (MCA) thrust will research and develop the non-extrudable components required by the grand challenge such as reinforcement, electrical, plumbing, and sensor systems and on the robots required both to assemble these components and to deploy the extrusion systems developed by the first thrust. The Integrated Software Systems (ISS) thrust will research and develop the software needed to go from design through construction, including planning and controlling the behavior of the multitude of robots to be developed by the previous thrust, and providing the logistical support required for constructing a house in a day, or beyond this to constructing a full community in a small number of days.

The Fundamental Research plane is home to four areas of fundamental research that map onto the needs of the thrusts. Materials and structures concerns the development and understanding of new composite materials with desirable properties, the flow of these materials, and the testing of structures built from these materials. Sensing and acting concerns the dynamic modeling and control of large-scale flexible robot arms, embedding capabilities such as sensing and power into structures as they are fabricated, and real-time inspection of structures as they are fabricated via vision and other sensory modalities. Systems and processes concerns the novel design possibilities and limitations embodied in the CRAFT approach to fabrication, and the processes by which all of the human and robotic participants work together from design through fabrication while being supplied with the resources they need as they need them. Modeling and reasoning concerns developing and understanding mathematical models of the fabrication process and environment, reasoning about fabrication geometries to determine feasible strategies for fabrication, and modeling and visualizing the whole process in both space and time. Results and enabling technologies arising out of these research efforts will be incorporated into the (sub)systems developed by the thrusts and will guide the development of the integrated engineered system that will be employed in the grand challenge testbed.

In addition to the research on the technology required for the grand challenge, research will also be necessary in several areas on the societal impact of the grand challenge:

Economic: Two concurrent “shocks” to the economic system are likely to occur: cost reduction of new housing and changes in employment opportunities within the construction sector based on newly required skill sets. Regional, metropolitan and sub-metropolitan economic modeling will be performed to identify macroeconomic impacts of these changes.

Employment: Strong resistance can be expected from existing construction firms and labor unions. Statistical analysis will be used to predict the probable employment impacts of the new technology. Analysis will focus on the extent to which the new technology will complement or substitute for existing labor.

Regulatory: Regulatory Agencies (e.g., Design Codes and City Building Departments) move very slowly to accept new ideas and methods. We will develop a testing plan for a provision of the code that allows for the use of alternative materials, design and construction techniques based on proven engineering knowledge.

Environment and safety: The construction industry has produced a disproportionate amount of environmental damage and personal injuries relative to other industries in this country. We will analyze the potential environmental impact of CC in terms of total life-cycle. To accomplish this task, we will design new sustainability indicators for each CC material flow in the economy at specific market penetrations. Safety issues will be compared with traditional construction by calculating the risks associated with related hazardous tasks.

Architectural: CC will provide new opportunities for creative design of housing. CAD modeling and CC provide a natural collaboration between design and fabrication systems for architects. CAD systems must incorporate CC construction rules to guide architects in designing feasible structures.

Development Phases

Figure3 shows the major milestones for the grand challenge and technology thrusts.

The Grand Challenge testbed of CRAFT is the construction of various housing structures, starting with construction of an average single residence house in one day, and advancing into larger scale and more complicated buildings. This section briefly introduces the three major phases of development.

Phase I. In the first phase the basic CC technology will be developed for automated construction of single-residence structures where a gantry system carrying the CC nozzle and other robotic arms moves on two parallel lanes installed at the construction site (Figure 4). A single house or a colony of houses could easily be constructed in a single run. Integration of the CC machine with a support-beam picking and positioning arm can produce conventional structures, while organic form (e.g., adobe) structures can be built without external support elements by using shape features such as domes and vaults. In this phase, we also plan to develop and integrate automatic embedding of reinforcement, plumbing, electrical and communication network wiring and sensor modules, as well as automated painting (using inkjet technology for wallpaper designs). While we will initially use the technology for emergency shelters and low-income housing in underdeveloped countries (Mexico, with the demand for nearly 500,000 houses per year, seems to be a good starting choice for implementation), almost immediately after its development we will address local building codes for commercial deployment of CC in the US.

Because of the unprecedented speed of CC construction, attendant improvements in the construction inspection process will be required. Therefore, we plan to develop advanced sensory systems and information technologies for automated real-time inspection and feedback to municipal computers overseeing ongoing CC construction activities at various locations.

Creation of the necessary information technologies for construction project management for the new technology is crucial. Material procurement is especially important for CC, as all needed materials must become available at the site in a short period of time. We will develop the needed project management software, communications systems, and supply-chain logistics for effective implementation of the technology on a major scale.

Phase II. In this phase the CC technology will be developed for construction of larger community and multi-residence structures. For apartment buildings, hospitals, schools and government buildings, for example, the overhead gantry platform (shown in yellow in Figure 5) can be extended above the width of the structures and equipped with multiple cross members, each holding a CC nozzle assembly and a robotic manipulator (for beam installation, plumbing, etc.). The integrated construction system in Phase II will include automatic methods for tiling.

Phase III. In this phase we will pursue the adaptation of CC for the construction of entire communities, including residential, public, and commercial buildings, public art pieces, as well as infrastructure such as roads, pavement, landscaping, water reservoirs, etc. We will also explore innovative designs, such as fractal communities.

We also expect to work closely with faculty at other universities and to add additional faculty at USC as things develop further.

Space and Equipment

In addition to lab and office space already assigned to CRAFT faculty and CRAFT related laboratories (Rapid Prototyping lab, Manufacturing Systems lab, Robotics and Control labs, Structural Testing lab, and Materials Science loabs) on campus, USC’s Information Sciences Institute has provided an office for the director plus a 900ft2 lab space for CRAFT at its Marina del Rey location. In addition, we expect that dedicated laboratories to be established at the site of most industrial affiliates of the Center.

Industrial Collaboration and Technology Transfer

CRAFT has developed an industrial membership program to extend its lead in automated construction technology and to transfer same to as broad a range of industries as possible. CRAFT also expects to leverage the experience of industrial members in developing solutions that contribute to overcoming the institutional and commercial barriers that always exist when migrating to a new technological paradigm. Industrial participation will be essential to the rapid transfer of technology. Finally, CRAFT expects to benefit from industry contributions, which will enlarge the research agenda and provide for exploratory efforts that may not get funded under core program funds.