Talking Shop: Professor Sanjay E. Sarma

The Small Shop Becomes a Virtual Factory


Sanjay Sarma

Professor Sanjay Sarma (Photo credit: Tony Pulsone)

Professor Sanjay Sarma has been a professor in the Department of Mechanical Engineering since 1996, after earning his PhD from University of California at Berkeley. Renowned for his co-development of standards-based radio frequency identification (RFID) technology use in commercial supply chain management, Sarma’s research focus is on efficient manufacturing and automatic identification, including barcodes and RFID,  supply chain management, transportation, and sensors. A member of the Laboratory for Manufacturing and Productivity (LMP), he is also the chairman of EPCglobal and is currently the director of the MIT/Singapore University of Technology and Design (SUTD) Collaboration.

Can you tell us about the current research you’ve been working on?

I am working on the development of a virtual factory. The idea is to enable small shops to present a united front and act as one big shop. It is like an ant farm. Every ant does one thing, but the ant colony is one big, breathing, living entity. In order to do that, you need to seize web-based technologies to bring the shops together and enable them to coalesce on the fly. We also need a better supply chain.

What are your ideas about how to connect these small shops and make them one?

There are three steps: distributed manufacturing, distributed quality control, and distributed shipping, all of which my team and I are looking at now. Distributed manufacturing is about getting independent machine shops to all do their part for the benefit of the whole. For example, if I have a small shop and would not normally be able to bid on a big project, I could work together with a couple other small shops to contribute a part of a bigger order. That is one piece of it, and I have worked on it for a long time. I am trying to make that more common. The second piece of it is quality control. Wouldn’t it be great if a machine shop owner could do a CAT scan on the parts I made? The reason companies like to do something in a single big factory is because “they only have one throat to choke,” meaning they can control quality and have accountability. But if the work is distributed, quality control is more difficult. But what if we used 3D scanning? Before a machine shop sends the part, it could send a complex and detailed photo of the part that far exceeds the awareness you would have in a factory. The third piece is now that I have made the part and it weighs 30 pounds, how do I ship it? UPS? At $100, that is way too expensive. You could put everything on a big truck to ship it off, but not if I’m only making 10 parts. So we are trying to evolve a new approach to supply chain planning in which small shops can rent unused space in someone else’s truck. It is called fractional ownership. RFID and other technologies will enable a new world in which the supply chain works as if it were a packet-switched computer network. The irony is that computer networks were inspired by the supply chain—or how it was thought to work!

How would this kind of setup affect factory employment? I think education is huge.

The United States is very computer savvy, so I have few concerns that we will innovate, but I do think we need to change our game. A lot of the unskilled jobs available will decline, but the great thing about America is how adaptive we are. And education is the key to all of this. I think there will be some job retraining, but I also think there will be some self-driven people learning new skills. A lot of people are quick to declare that America is in decline, but you see more spark of entrepreneurship in an average person here than you sometimes do in business leaders in other countries. It is an amazing thing, and you have got to trust that. But you have also got to facilitate that creativity. Distributed manufacturing may be one such facilitator.meche logo