I had the opportunity to attend the CCC-sponsored workshop, “A Research Roadmap for Robotics in Manufacturing and Automation“, which took place in Washington, DC on June 17, 2008. Below is a loosely-edited excerpt of the notes I took during the workshop. The intention is to convey a general sense of what happened at this meeting, and how we can apply the lessons of this workshop to other CCC initiatives.

Workshop Notes (excerpts)

There were 35 people in attendance, including Joe Bordogna (former COO NSF), Clint Kelly (formerly DARPA), Elena Messina (NIST), William Joyner (Semiconductor Research Corporation), people from industry (General Motors, General Electric, ABB, C&S Whole Grocers, Willow Garage,…), plus academics (GATech, CMU, Berkeley, Utah, Colorado, UPenn,…).

This workshop was unlike those that typically happen at research conferences.  The discussion was not about pure science, but the intersection of science, national needs, public policy and funding.  There were almost no prepared talks. Instead, the program consisted mostly of group discussions, break-out sessions, and consolidation discussions with all attendees together.

Henrik Christensen gave a brief initial presentation to set stage. This was then followed a series of presentation by non-academics:

• Joe Bordogna: On how big science gets funded.
• William Joyner: On collaboration between industry and research in the semi-conductor industry.
• A series of talks by industry folks (GM, GE, food industry…). What is the state of the art, and what is needed? These presentations made clear the large impact of robotics on national infrastructure and economy, what past/current techniques are, and what new technology is needed for progress.

After the presentation, the workshop divided into breakout groups:

• Societal/Business drivers for robotics. Why does the country need robotics?
  • external drivers: inflation, human-resource costs, energy, environment
  • demographics: aging workforce, different skills/job expectations
  • manufacturing as a critical technology for economy and security
  • maintenance/management of national infrastructure: bridge painting…
  • successful design (our current strength) requires manuf. know-how
  • trend toward personal manufacturing: lot-size=1, customized products
  • traceability: salmonella, tomato processing…
• Obstacles to progress in these drivers. What research/technical progress needed.
  • Henrik: “Only at level that could be understood by Congressmen.”
• If given $$$ for robotics, what would you invest it in? Included technical discussion in both application and theory.
• If you had to write a roadmap now, how would you tell the story. How would investing in robotics make a difference?

By bringing together academic researchers, industry experts and public-policy experts, there were great exchanges and discussions that don’t happen at typical research conferences.

After the workshop, a number of documents were produced. That night, immediately after the workshop, a “DRAFT DRAFT” version of workshop report was produced. Later, a draft outline of of “roadmap” for robotics in a manufacturing and automation was also generated. This was presented to a Robotics Congressional Caucus, which involved congresspersons from Pennsylvania and Tennessee. (The documents will be made available on the web at www.us-robotics.us/blog. (Register to be able to comment on the report and participate in community discussion.)

The plan is to have a revision and synthesis of comments by September, in time for a possible review meeting at IROS’08, and then a synthesis workshop around November 2008. The goal, then, is to have a first complete document in November 2008, and presentation to university presidents and others in December. (There was a suggestion to organize an open meeting at IROS to get more feedback from the robotics community. However, IROS organizers said that it would be difficult to make room for such a discussion. Other possible conference venues are being explored.)

At one point during the workshop, I asked Henrik Christensen several questions:

Andrew McCallum: How did Robotics Congressional Caucus start?

Henrik Christensen: The President of CMU started it on his own initiative. The CMU Robotics Institute is important to CMU, and he wants to keep it strong.  (Side comment:  University Presidents are a great, highly-connected resource, and we should think about ways to leverage them more often than we have historically.)

MC: What is the end game for your CCC robotics initiative?

HC: I’d like to make menu of research opportunities, in the form of a 2×2 matrix. On one axis we have long-term vs short-term, and on the other axis we would have applied research vs basic research. We could then take this to many congresspeople, pitch lots of ideas, with the hope that some of them get excited. Then, we would ask for their help in enhancing robotics research opportunities. The entries in the matrix could also be targeted to specific funding agencies:

• short-term applied: go to NIH
• long-term basic-research: go to NSF
• long-term applied: go DARPA,…

Concluding thoughts

I think the organizers of this effort are on the right track. They have their sights focused on things that aren’t already provided by existing workshop venues.  The set of participants and the break-out sessions were all very high quality.  To his credit, Henrik runs a tight ship.

The CCC should encourage and find ways to provide even more support for this effort. In particular, Henrik and others are trying to make important governmental connections on their own. The CCC Council should be well equipped to help him. Even more importantly, the CCC should provide services, pointers, and contacts to other CCC groups that aren’t already as savvy as Henrik.

I thought there was great value from inviting some people outside robotics, in neighboring fields. The CCC should encourage other workshops to do this. We should also encourage other CCC initiatives to structure their workshops like this.

— Andrew McCallum