Thanks for taking time to comment. I'm definitely interested in understanding where you see a disconnect. My understanding of commercial off-the-shelf (COTS) has always been just as the name suggests – they constitute a standard, commercial product line designed to meet a standard set of specifications for applications across a broad range of markets. A COTS linear regulator, for example, might be used in architectural lighting, consumer electronics, and automotive designs. The broad customer base generates volumes that keep prices down. The devices may not literally be available off the shelf, but they are rapidly available. I consider commodity components to be a subset of COTS components. Commodity components are economical, available in volume, and basically vendor independent. COTS components are economical, available in volume, but are not necessarily as interchangeable.
Does that fit with your understanding? If no, how does your definition differ?
The problem I was trying to highlight in this blog, which I have heard about from the customer and vendor side over and over, is that consumer markets move quickly, and carry manufacturing with them. If a COTS component has been aimed at a specific market sector such as cell phones or digital cameras, and now that demand has evaporated because the latest generation of those devices imposes new requirements, the COTS component vendor will most likely move on and begin producing new products for their volume market. As far as they are concerned, in the context of their customer base, the older component has become obsolete. Continuing to produce it is no longer cost effective. In a perfect world, a similar device might be available from other vendors, but there's no guarantee that they will satisfy all of the specifications required for, say, a replacement MCU deployed in a Humvee built two years ago.
Do you agree? Disagree? Please drop me a line and let me know.
As an engineer with twenty years experience in military avionics power supply design I whole-heartedly agree with the author. I also know well the basic concept of COTS. Most program managers pick a COTS solution to avoid the NRE on a custom design. Then once it gets into the field, they end up paying the price when the part, such as a COTS power supply, doesn't work and no one knows anything it. The company that they bought it from either no longer exists, wants a lot of money to fix it, or they don't know how to fix it either. I have been down that road before. Custom designs in-house cost more upfront, but they save a lot of money over the life of the program, bank it.
It all depends on your tolerance for failure.
It also depends on what the the contract allows. Always check there first. Than, before you commit anything to your CAD/Drafting department, TALK TO YOUR FRIENDLY NEIGHBORHOOD COMPONENT ENGINEER. It is his/her job to keep your parts selection within the parameters set by the program.
P.s. I have a lot more to say on the subject. A lot more, but I would be exceeding the limits set by EETimes. So this is my "tweet" response
I wish the article had explained eliminating lead from solder used in electronics equipment is a European Union requirement under RoHS that has no counterpart in the U.S. Unfortunately, the leads or conductive portions of most component parts are now tinned in almost pure tin - from which grow tin whiskers that cause shorts.
A Book For All Reasons Bernard Cole3 comments Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...