[Guest Blogger] 40 Years in the Field (Part 8)

Graham Jones (Surrey, British Columbia, IT Pro)

Part 8 now finds me in Canada. So what brought me to Canada apart from a ride on a “paraffin budgie”? Well, I was out of work for one thing but I heard through my good friend (the one who used to be my boss back in part 2) that an engineering company in Vancouver was looking for someone to start up the use of Computer Aided Design (CAD) from scratch. I had done my “practical” engineering bit and computers were beckoning again. By sheer coincidence CAD is something that I had started to take an interest in and the idea of starting from a “clean sheet” was appealing, and so was the prospect of doing it in Vancouver :).

When I say a “clean sheet” that wasn’t completely true because a study had already been completed into software packages for doing 3D plant design. The package that had been chosen (well it kind of picked itself as you will see) was Plant Design Management System (PDMS). It had evolved out of research work done at Cambridge University in England and was spun off as the CADCentre. What I haven’t told you is that the company in Vancouver was a subsidiary of CIL which in turn was a subsidiary of ICI, where I had been working. ICI had chosen PDMS and so free training and some support was on offer. So in actual fact I came to Canada and then went back to England (ironically to where I had worked in part 2) for 4 months on the job training. PDMS is now marketed via a company called Aveva.

Of course, there were other 3D packages available but PDMS was a “different breed”. Other packages had evolved as extensions of 2D drafting and permitted you to draw in 3D (and associate data with the drawn objects) but it was still drawing and the “model” effectively existed as electronic sheets of paper. PDMS tackled the problem completely differently. You literally build a 3D representation of the plant using basic 3D shapes (objects) such as boxes, cylinders, cones, rectangular and circular toruses (or should that be “tori”?), etc. positioned in 3D space. Collections of shapes could be grouped to form a piece of plant equipment and located/moved as a whole. This was all stored in a DB called ADABAS. In addition to dimensional and spatial data you could store properties which could be used for reporting purposes. For example, you might store the material of construction. You could also build libraries of components to select from. This was particularly important for building piping systems and steel structures. Drawings were still important because that was the working communication medium both internally and externally. The drawings were generated from the DB. Having built a “model” it was possible to produce a limitless number of different isometric and perspective 3D drawings. The Project Manager’s (PM) view was supposedly from underground :).

When I got there they had nothing, no computer hardware, no software and no money; well, very little money (:. So it really was “from scratch”. The other thing that soon became obvious was that PDMS was like an Xmas toy, “some assembly required, batteries not included”. It was going to take a lot of effort building piping and steelwork libraries before you could do anything useful. You could buy them but we didn’t have the money! The software was very expensive (approx. $400K for a license plus an annual maintenance fee of $10K per user in today’s dollars – software pricing has changed somewhat since then). The only choice was to use a local computer bureau (incidentally, PDMS only ran on Prime minicomputers at that time – more of that later) who could rent the software on a per use basis from the vendor. So armed with a dial-up connection and one dumb terminal with a monster 12” monochrome screen we “launched” with very little fanfare (:. Ever had that sinking feeling that something is going to be a much bigger battle than you thought? I am of course referring to the people component again. Everybody was quite happy with their pencils and electric erasers drawing on vellum and nobody (ie. the company management) was insisting otherwise. Sigh!

So it was on to plan “B”. If you can’t beat them, infiltrate their ranks. It was clear that since nobody was going to listen to “little old” me I had to find champions from within the designers who could “lead the rest to the promised digital land”. In other words, I needed some “disciples”. Luckily, I found 2 easy convertees (the beatings did continue…) one of whom eventually came to work for me and turned out to be a computer and programming whiz (he had been taking night classes). I had come along at the right time to free him from his mundane job as an instrument draftsperson (hail Caesar). The other person was a very senior designer with a progressive attitude who could “see the light”. In other words he understood the opportunity that PDMS represented. The rest could only “see” paper!

The next challenge was to find a PM who was willing to “give it a try”. Will all those volunteers take one step forward? Where did they all go? When I first got the job I was asked how long it would take to achieve a reasonable measure of success. I, of course, said “at least 2 years”. 2 -3 years is the standard answer BTW regardless of how long you actually think because if you say 4 -5 years nobody ever does anything! I won’t hold you in suspense. It actually took about 7 – 10 years depending upon where you judge the “end point” to be. Of course you never really reach the end point because the “goal posts” keep moving. Although there were technology changes (software and hardware improvements) in that time the main obstacle was the people, and it usually is! You can lead a horse to water but you can’t make it drink. Since I had no PM volunteers we were now onto plan “C”. The sales and marketing people usually have some “discretionary” money. Perhaps we could wow our prospective customers by including “sexy” drawings in the proposal packages, ie. give them a “preview” of what their plant might look like. Further, perhaps they might insist that we use PDMS on their project having seen what it can do. Dastardly cunning! Even more amazing was that my tactic actually worked. Such “see before you buy” activities using computers are commonplace today but back then it was fairly novel.

Why did it take it take so long? I have already mentioned the “people” factor but what was really going on was a paradigm shift in the approach to process plant design and engineering. In other words it was a new era of Computer Aided Engineering (CAE). What is the difference between CAD and CAE. CAD is a simply a “tool” whereas CAE is bound up in the whole process (from cradle to grave) of process plant design, engineering, construction and operation. What I had realized several years earlier is that the whole industry would change when it recognized that “it was all about the application of traditional data processing techniques” and not engineering per se. Why not? There are masses of “data” involved in a process plant project and “data flow and concurrency” was a huge challenge, and still is. Up until this point data processing was regarded as what you did with computers in the “commercial” world. So I had taken my new job with that realization and intention in mind. My employer hadn’t the faintest clue about my “longer term plan”. Sometimes you have to drag people “kicking and screaming” to get them to a “better place”. Being a “prophet in your own time” certainly isn’t for the faint of heart. In part 9 you will see how we evolved from the very humble beginnings and something of the start of our “march” (actually it was more like a slow “crawl” on our hands and knees at times) from CAD to CAE. It was a challenging but interesting and rewarding “journey”.

Cheers

Graham J.