William Strecker, the then CTO of Compaq, writing in the December 1999 issue of PCQuest, talked of a future when the computer would simply disappear from sight. Without calling them so, he was talking about embedded systems. Reading that piece then, I thought the concept to be interesting; something that would happen at some far away date.
These days we talk much more of embedded systems. But then, all we talk of are tiny systems, embedded into equipment like refrigerators and lifts, and the like. The systems that control or monitor the operations of these equipment is what concerns us. And even then, it seems to be mostly talk and less of action.
Fast-forward to last month when we were researching this month’s story on technology in medical systems. That was when the whole concept and scope of embedded-computing systems really struck me, in all its vastness.
Take the case of a CAT scanner. CAT stands for Computerized Axial Tomography. So, it is fairly obvious that there is a computer inside it somewhere. What it does is also fairly easy to deduce. It helps assemble the individual scans into 2D or 3D images. But what are the specs of that computer? How much memory? How large a hard disk? What graphics cards? What OS does it run? What applications?
We hunted high and low for this information. Doctors and medical books had no clues. Obviously, they were not concerned with what made the CAT scanner work. Internet searches threw up nothing more than vague hints or extremely in-depth theses on specific algorithms.
And finally, when we got the information, we were in for a cultural shock of sorts. Instead of the miniscule embedded systems that we had expected to find, what we came up with was full-fledged workstation-class machines inside these scanners! Yes, workstations running various versions of Unices or Windows. And in one case we came across two workstations inside one scanner!
It was a completely different perspective to embedded systems, at least for me. Just think of it–a full workstation as an embedded system!
A quick search indicated that what is true of medicine and healthcare is equally true of space exploration and telecom and entertainment and manufacturing and many, many other sectors. The need for computing power in whatever they do, or whatever they create–from the car assembly lines to the cars themselves–is
a given.
It is my hypothesis that fairly soon, the demand for creating software and hardware for such systems could overtake the demand for creating good old PCs and servers and software for them.