My career path started with experimental high energy particle physics as an academic pursuit, and then changed to the application of machine vision to manufacturing in the outside, or industrial, world.
My interest in physics had been triggered by a fascination with the fundamental workings of nature; performing basic research on elementary particles appeared to be a good way to satisfy that. But acquiring the Ph.D. in a world-class environment (Columbia in the 1960s) set high standards. And alas, over the years, with long experiments to answer not-very-fundamental questions, and with large and ever-growing collaborations (there are experiments now with over 600 people!), physics lost some of its luster for me.
An opportunity to leave academia came along, and soon after another one to join the emerging machine vision technology at Automatix Inc., a start-up company in robotics and machine vision. That's the field in which I've spent the last twenty years. But not at the same company; that just doesn't happen in the high-tech industry.
(See below: Life in High-Tech ).
Along the way, through the ups and downs in a number of companies, I saw and experienced a lot of work conditions, and learned to deal with and adjust to most of them.
(For more, see Learnings ).
(For more, see Poor Management ).
(For more, see More Management Learnings ). This includes some examples of good management.
The variety and challenge in machine vision is what makes it so fascinating; to illustrate some of that, here is my last resumé , from August 1998 (no, I'm not looking for another job!).
The last vision product that I worked on, in Teradyne's Intelligent Imaging Systems group, was the Optima 7200, an inspection system for circuit board assembly. My responsibilities included lighting, optics, camera/lens/lighting mounting, calibration and alignment of the camera/lens assembly, and characterization of lens distortion, all in the framework of color vision and high precision measurement. So it was with some pride that I would look at the camera/lens assembly on the pictures in the product literature. Then on April 22, 2003, this assembly and its capability for color discrimination was granted Patent No. 6,552,783: to see pdf version, click here. Now that I don't need it for a resumé any more, I have become a co-inventor with a patent!
Invariably, when people ask what I do (that is - what I did) and I say "machine vision engineering", then I need to explain just what machine vision is.
Basically, in machine vision, images are captured by a digital video camera and then analyzed by computer. For industrial application (as distinguished from academic research), the purpose of this analysis is inspection or robot-guidance. Inspection, in this context, means checking that the part being viewed is not missing something (defect detection) or that it has the right dimensions (gauging).
The challenge in industrial applications is to perform this inspection or robot-guidance for assembly-line production on the manufacturing factory floor, which means that it has to be inexpensive, fast, precise and reliable.
The manufacturing floor also has its fascinations for those open to them (which I was). This is where the products that we, as consumers, are assembled and tested, usually at high speed and high volume. The environment is often noisy, dirty, and greasy, but that's where the action is. You haven't done real programming until you've made some customer-requested changes in a non-airconditioned, gritty, noisy brakeshoe factory in Florida in August when the outside temperature is in the 90s.
The work in machine vision is challenging and exciting, but "security" is not in the High-Tech dictionary. On the other hand, security, as in academic tenure, can also be a golden anchor.
In fact, in High-Tech, even "stability" is usually temporary. To illustrate with my own work history:
Closing a division: When I left academia in Sept. 1979, it was to go to Union Carbide Imaging Systems, which was developing medical imaging equipment. Although Union Carbide was known for old technology (anti-freeze, batteries, etc.), this was to be their visible entry into new technology. It turned out that three months after I joined, the company closed the division.
Rise and decline: Fortunately, machine vision and robotics, as emerging technologies, did not yet have a lot of experts in the field. A start-up company, Automatix Inc., was looking for raw talent with the opportunity to become experts. The company had a dynamic leader (Phil Villers) who knew how to attract venture capital, so it grew rapidly and gained a lot of market presence. It was an exciting time with new technology and lots of smart people working together on challenging problems while learning about the intricacies of machine vision, robotics, welding, and the customers' manufacturing facilities. Unfortunately, the company also lost a lot of money and, after four years of growth without profit, went into slow decline. Six years after I started at Automatix, with an offer from Digital Equipment Corp. in hand, I left because I felt that it was time for new blood to help restore the company's fortunes. Since then, Automatix merged with Itran to become Acuity, which was then bought up by RVSI.
Downsizing: When I joined Digital Equipment Corp. (aka Digital, or DEC) in 1986, "downsizing" did not yet exist as a concept or as a word. The company was the leader in New England High-Tech, with the fabled Ken Olsen at the helm. Within a few years, the decline began and then continued inexorably as the company changed management, continually shed employees and closed facilities. We learned about "RIF" (reduction in force), "redundancy" (the overseas term), "right-sizing", "outsourcing", "layoffs" and "downsizing". My optimistic belief that things had to get better was proven wrong, until I finally also became a downsizee. Since then, the company kept declining until it lost its once-honored name in the acquisition by Compaq, which has in turn been acquired by Hewlett-Packard.
A step back to go forward: With a wish to go back to hands-on engineering, it made sense to lower salary expectations and look for a small company, in this case an integrator, IPC (Interactive Process Controls), looking to expand its machine vision capability. After two years, one of helping to build up machine vision and one of seeing it decline, it was time to change. I made the decision not to be a full-time employee any more, but to be available for consulting.
Contracting and Consulting: A five-month contracting opportunity on a challenging project came along; it was one of the most enjoyable projects yet, to develop alignment and calibration for a new piece of equipment, at MPM, one of the two leaders in solder paste deposition systems (MPM has since become part of Cookson Electronics, or perhaps Speedline Technologies.). At the same time, I was still doing some consulting for the integrator IPC. This consulting expanded then to the equivalent of full-time when the contracting project ended.
Product Cancellation: Five months later a contact from a recruiter led to a full-time position with GenRad, on a team developing vision inspection equipment for the assembly of circuit boards. IPC folded soon after. At GenRad, nine months later the product was canceled on short notice, with the vision engineers being let go just as quickly.
Pre-retirement: During the next few months, I pursued a couple of promising leads, but started getting used to the idea that, with retirement not too far in the future, there were lots of good uses for the free time away from work. One of the new activities was volunteer math tutoring at a nearby high school (Cambridge Rindge and Latin).
One last time: Then a call came from a group at Teradyne also developing vision inspection equipment for circuit board inspection. We agreed on a nominal 32-hour work week and an hourly rate, so that I could continue my tutoring two afternoons a week. New technology provided me the chance to do some learning along with contributing novel solutions, one of which led to a patent. Twenty months later I made my final decision to retire. Soon after, GenRad was acquired by Teradyne.
In summary, only the most recent company for which I worked continues to exist under its original name. All others merged, were acquired, or ceased operation.
Oct. 31, 2005 Update: Last week I met a former Teradyne manufacturing engineer who told me that the inspection product I had worked on had been canceled a few years before and that the development group was disbanded. Sure enough, when I looked at the Teradyne website, the Optima 7200 was no longer listed as a product, and the only references were in some press releases and other old documents.
At the time I retired and left the development group, the product had just gone out for evaluation and testing at some selected customers. Typically, this is when the gap between customer needs and product performance becomes evident, and the difficult work really begins. Circuit board assembly inspection with a vision system is a very tough problem because of the great uncontrolled variation in visual appearance and color of the boards and components. In my opinion the development group underestimated the difficulty, because it had not only technical and financial, but also a great deal of emotional investment in the machine. My suspicion is that because of this the system never worked as promised, and that during the high-tech downturn Teradyne decided to cut its losses and return to its core businesses.
Fortunately, I had downloaded the pdf file of the product literature, and so can still provide access here (Optima 7200).
That was life in High-Tech!