Advanced Manufacturing Government Panel

December 8, 2011, Advanced Manufacturing Partnership, UCB, Berkeley, CA—A panel of government agencies discussed the issues for next-generation manufacturing. Arati Prabhakar from the US Department of Energy moderated the panel. Panel members included David Brinkley from the Department of Defense, Leo Chrisodoulou from the Department of Energy, Patrick Gallagher from the National Institute of Standards and Technology [NIST] , Bruce Kramer from the National Science Foundation, and Thomas Lee from DARPA.

Examples of your work?

Leo Chrisodoulou offered the parallel lesson from the 19th century. Aluminum was a very expensive metal because the extraction process was difficult and expensive. When electric smelting was invented, it changed the technology for extraction and the cost structure dropped to almost nothing. This example shows what they are trying to do. They are investing in basic processes to transform the way to make things.

Silicon Valley is one center of change. One new process technology is in focal plane arrays. Today, these are still a fire and brimstone manufacturing process. They are looking into low temperature manufacturing possibilities, because over 60 percent of all energy goes to thermal processes. Their efforts are directed towards energy efficiency and renewable energy.

Patrick Gallagher focused on how to implement new processes. They help to fund shared facilities with collections of expensive capacity to enable early experiments. One example is the nano-science center. It has a collection of high-end manufacturing tools and a collaborative R&D environment to advance new materials.

This is a model for other areas. Sematech shows how to share technology to develop a replacement for CMOS. The nano-research initiative will explore new platforms through more consortia. These research groups will create a commons to share resources and data.

Thomas Lee stated the themes for the 21st century will be synthesis. This technology will encompass electronic, biological, pharmaceutical, and material developments. One challenge is to scale the activities from their boutique and hand-carried modes to mass production. They need more ideas to try out. More people working on a subject leads to more ideas and some will become better solutions for the manufacturing areas. One issue is that decoupling manufacturing and design is only good in some areas. Other areas, like living foundries, need to share the information and technologies.

David Brinkley is interested in acquisitions, technology, and the ability to change logistics. With a $200B procurement budget, the DoD needs industry partnerships to improve program management and advance manufacturing processes. They need to identify advanced manufacturing capabilities and move them into DoD contractors to reduce costs and development times. Title 3 is an effort to make technologies more mature and commercial ready. It these technologies are ready, they can move to full manufacturing with some investments.

Bruce Kramer noted that they fund fundamental R&D. One example is the Google library project. Another is the idea of IC design rules that came from a Stanford senior design course. They look at both long- and short-term horizons for their efforts, as evidenced by their work in synthetic biology and the research center at Berkeley. In the area of nano-manufacturing, they are looking at plasmonic imaging. They want to take ideas to the point where industry gets interested in furthering developments.

Specific manufacturing capacities in the next 10 – 15 years?

The Department of Energy considered low cost carbon fiber. Now, all applications are for high performance applications because of the costs and difficulty in manufacturing at high temperature. A cost of less than $5 per pound would decrease costs and increase design flexibility. The industry has to reduce the cost of the precursors and democratize carbon fiber to make it like aluminum. We need a spectrum of fibers for greater diversity of opportunities.

Other areas include education, chemistry, people, and equipment. Other materials properties need to be defect driven, because most other manufacturing is not like silicon. Designers must care about the extremes and use probability for all parameters. New materials have to define minimum specifications for properties to make it easier to design.

NIST expressed a needs driven area. Manufacturing excellence drives innovations in materials, processes, and technologies. It serves a national need and is one of the engines for the economy. The deep impact innovations will be in nano, information, and biology. Life sciences will disrupt the whole economy. Biological manufacturing will be a big challenge for regulators to ensure safety and efficacy. The biological area will change many industries and the government.

Thomas Lee is concerned about the coming brain drain as the baby boomer generation moves into their 70’s. This transition will put a huge load on healthcare and government. As a result, healthcare will have to change from reactive to preventative. This change will require many more sensors that can detect and track symptoms. The costs of the sensors will have to drop and health systems will have to address the issue of real-time monitoring.

Decoupling design and manufacturing?

Patrick Gallagher opined that decoupling in the semiconductor industry has been good. However, decoupling is not the same as innovation. There are too many boundaries that reduce the capacity to innovate. CMU is investigating the issue of manufacturing versus R&D. optimization and automation have short term costs, but the benefits affect new products almost 10 years out.

Both coupled and decoupled design flows are correct, but manufacturing is a series of systems that require interoperability and supply chain management. Manufacturing agnostic design leads to functional segregation and generates more mistakes. An example is the latest Boeing airliner, where external suppliers designed components without understanding their relationship to other parts. New developments need interaction and the transfer of information across boundaries.

Bruce Kramer considered this to be an area of tradeoffs. The issue is complexity versus efficiency. Automobiles are not like semiconductors. Where do the opportunities lay? In the semiconductor industry, Moore’s law is the driver for change. Other industries don’t have similar drivers.

Thomas Lee suggested that this is an issue of control for time and is driven by increases in changes. Decoupling allows for lower efficiency in some areas to be compensated by higher efficiency in others. The future of manufacturing objects is in the long-tail markets, those small volumes that aggregate to large dollars. The future will be in tailored objects that may not need any existing manufacturing capacity.

For shared facilities, is there a national catalogue of functions?

Patrick Gallagher answered there are some shared resources. Some commons exist and information is shared where people work together. Everything happens at the boundaries of manufacturing and design. These centers of excellence are also safe havens where industry, government, and academia network. They share assets, databases, and create catalogues. The networking functions are what make partnerships work.

Leo Chrisodoulou responded that they track manufacturing demonstration facilities, incubator models, and collections of equipment. The listings will be available on-line soon. They are looking for space for manufacturing development, especially to large scales.

Does the technology exist for integrated systems that drop down to the employee level?

Leo Chrisodoulou suggested that we need to enable more cottage industries. An example is the smart phone. It allows mass customization through apps, rather than make products in small quantities. However, we still need cost-effective, low-volume distributed manufacturing.

Bruce Kramer commented that this is a fundamental issue with the Web and infrastructure. The capabilities exist for small businesses, which didn’t need the government to develop and support any development.

Patrick Gallagher suggested the workforce needs to change. The image of workers in small manufacturing is no longer that of a guy standing at a machine, but is now a high-skill job requiring education and cross-platform knowledge.

Is the workforce ready, is the education system up to the task?

Patrick Gallagher stated that the agencies and all other organizations are working on this issue. Not only are high-end universities involved, but also work has to be done at the high school and associate degree levels. All are creating programs that focus on STEM (science, technology, engineering, mathematics). NIST is pushing for curricula changes to address the new needs for educated jobs. Unfortunately, teachers generally have no experience with the jobs, so industry has to get involved with education.

The role of the national labs?

Patrick Gallagher responded that they have existing capacity of resources and are becoming increasingly collaborative (except for those areas requiring security clearances.

Leo Chrisodoulou noted that they are working to reduce friction and make it easier to get access. Now there is a standard interface process.