Category: Computer Analyst

Shortly after moving to Toronto I took a job at Litton Systems Canada (LSL) — a division of Litton Industries. This division was primarily tasked with producing guidance systems, and at one point most of the commercial airline guidance systems in the US and Canada came from LSL.

My position was in a section called Test Equipment Engineering. They were responsible for making sure that all pieces of test equipment used in manufacturing were precisely in spec and accounted for. There were literally thousands of items that were considered “test equipment” — anything from a sophisticated electronic device down to a screwdriver, and also the software that ran any automated tests. Our task was to track every item, recall it according to its schedule, have it tested in a lab, and then update the records and return it to use. Our customers (many contracts were from the US pentagon and other countries) required such records be kept accurately.

When I started there, they had a rudimentary program based on old hierarchical database technology, but they wanted to modernize and improve the system. I was in charge of the migration to Unix-based computers (Hewlett-Packard HP-UX) and a relational database (Informix). I was instrumental in selecting the computer model and other related hardware, doing most of the database programming, and managing a few other workers.

The system pumped out printed forms for each item that was due for calibration, then these forms went to the calibration labs where engineers would retrieve the physical devices (each had a serial number tag). After they were calibrated (tested against standards), the forms were filled in and the devices were returned to use. Some devices required calibration only every few months, while others called for weekly or bi-weekly testing.

I was also responsible for our department’s computer terminal installations and re-locations. In those days, workers had a terminal on their desks, not a computer. The terminals were connected using a COAX cable, and communicated directly to the UNIX computer. One of the attached photos shows the COAX cable tap that was used to connect an individual terminal to the coaxial network line.

Another project that LSL worked on was the first “heads-up” display for the Apache helicopter. Although I wasn’t an engineer, someone offered me a tiny role just for fun: The system had some control buttons with a novel LCD backing that could change icons depending on its function. I was given the opportunity to design one of the LCD icons! I don’t recall what it was, nor do I know if it was ever actually used in production, but it was something of a thrill.

The Litton job (which I kept for 12 years until we left Toronto), was my introduction to UNIX, Hewlett-Packard, and relational databases.

In 1992 we decided to move from Toronto to Chicago. After 12 years at Litton, this was a major change in my career. After taking a few months to settle in Chicago, I turned to job searching. I found an ad for a consulting position at Abbott Labs headquarters in north Chicago, and easily landed that temporary position.

Turned out that one Abbott division was trying to move from legacy mainframe computing to HP Unix-based systems to save money. After a few months working at Abbott, they offered to make me full-time and I accepted.

The division I worked for was called Clinical Services, and was responsible for gathering, analyzing and reporting on all the clinical trials data for new pharmaceutical products. Of course, my role was as an HP Unix expert, figuring out the hardware, software and design best suited to meet the division processing requirements. The work was interesting, and looked to be a 3-year project to get all the users and applications moved to Unix.

The Abbot Park location was beautiful, very much like a university campus, with multiple buildings in a large park-like setting. Many of the people I interacted with also reminded me of a university environment, since they were PhD medical researchers.

Working at Abbott gave me a deep insight into the trials and tribulations of pharmaceutical research. Although these companies charge what seems like crazy prices for new drugs, and they do make huge profits, then also take huge risks. A small anecdote that made me more sympathetic to this: There was a research scientist that I became friendly with. I’d pass his office a couple of times a week and often stop for a brief chat. He was always at his computer with all sorts of molecular models displayed. One day as I passed I saw him sitting in front of a blank screen. I asked why and he told me that after 5 years researching a promising new drug, he had just proven it to be ineffective! His last 5 years was for nothing, as he said “I’m back to the drawing board.”

I was also blown away by the requirement to print literally 3-5 CARTONS of paper reports about each drug trial — and the FDA required 3 copies, so that could be up to 15 cartons! I had to set up a room with multiple high-speed printers and design a method to organize the print queues to optimize print times while keeping reports organized.

I was two years into this 3-year project when I was spirited away by a local headhunter that I knew. We had already showed solid success by moving more than half of the users and applications to Unix, with the costs being less that a single year of support costs on the mainframe. Moreover, turn around times were faster, so everyone was happy. This headhunter described a new venture within a large company that desperately needed HP Unix expertise, and I had the perfect reputation. The company was Mercer, described in a separate post.

A headhunter I knew called me while I was at Abbott Labs, encouraging me to check out a new job. Although I was happy at Abbott, the challenge sounded interesting, so I went on the interview.

William Mercer, a large New York based insurance company was launching a new venture which they called the Mercer Administration Center (MAC). This was to be a service to outsource employee medical insurance management for large corporations. They had been at it for a year, but were facing some major challenges, many of them, it seemed, HP computer related.

The idea of being involved from the ground up in a new venture sounded good, and they wanted me to take over total ownership of the computing requirements, including showing them how to properly separate the phases of the software development lifecycle (Development, Test, Production.) In addition, they were offering a significant increase over what I was getting from Abbott, and the drive from home was less than half the distance. So I accepted.

After a few weeks I started to understand what had happened there. Seems that the “subject experts” — the ones responsible for figuring our all the ins and outs of a medical insurance management system — didn’t really understand the business. As a result, the company failed to deliver as promised to their first big customer, Pepsi. They had quoted a one year development time, but then failed to tell the customer until after 11 months that they were not going to meet their own deadline. When Pepsi asked how much longer they needed, they said “another year!” Remarkably, Pepsi agreed…

I did my job and established a proper software life cycle computing environment, and trained the programmers on how to properly use it. Over the months that followed, I became very uneasy because it was apparent to me when attending business meetings that they still seemed to have the same problems. Soon, it became clear that one upper level manager based in Louisville was actually lying about the project to the Mercer Insurance parent company! He was (again) re-assuring management that they were on track, when in fact everyone knew they were a mess.

I could clearly see that they were going to miss delivery, probably by another year! So I jumped in (because I’m not the type to silently watch something like that), and at a company meeting where that manager and others were re-assuring all the staff that everything was on schedule, I stood up and asked how they expected us to believe that when we could see that things were still a mess. There was a bit of a tense exchange, and I stated that Pepsi was highly unlikely to grant us another year to complete the project. The meeting, needless to say, became quite noisy, and was quickly ended. Afterwards I was approached by many employees, thanking me for having the guts to state publicly what we all knew was true.

However, I could see the writing on the wall, so I called up my friendly headhunter and said I was ready to move on. Luckily, he found me a great position — because a few months later, William Mercer found out the truth and pulled the plug on the MAC, shutting it down permanently.