Earlier in my career, I worked on a project that integrated a robotic body-welding line for Ford Motor Company. In those days, Ford used Kawasaki robots almost exclusively. They had a fixed hardware and software interface standard for each type of robot with a long and storied history.
My project began with an attempt to sell Ford on using a different, better (but more expensive) brand of robot, and a better (but again, more expensive) pinch-welding "gun." This process proceeded apace until about the 11th hour, when suddenly some intransigence from one particular individual at Ford suddenly forced everyone to drop the original plan, cancel the original hardware orders, and order Kawasaki robots. (The better weld guns, however, were kept.)
As the designated shop-floor "robot guy" for this project, I became involved after all the hardware components had arrived, had been uncrated, and were being assembled. Innocent of the consequences of this project's sordid history, I sailed blithely into the mess and set about putting together the various bits and pieces, as I had done many times already in my short career.
It didn't take very long to discover a serious problem.
Back then, the very old, set-in-stone, Ford/Kawasaki standard used a single multi-conductor cable to carry discrete I/O signals (24V DC and even, in some cases 120V AC) from the robot cabinet, through the weld controller, and out to the weld gun. For those of you not familiar with the term, a weld "gun" is the "pinching" bit that you see in movies like Minority Report, or any PBS special about the automotive industry. We encountered our problem when we attempted to wire up the new weld guns to the old cable.
The old cable provided only two signal wires for closing the pneumatic valves that operated the weld gun. The new weld gun required four wires. And the kicker was that although the Ford/Kawasaki standard had provided for several spare wires in its youth, feature creep over the ensuing years had whittled the number of spare wires down to one.
Before calling upstairs to the design engineers responsible for this strange turn of events, I pulled all the electrical drawings: the gun, the weld controller, and the Ford/Kawasaki interface that the Kawasaki robots had been delivered with. I discovered that they didn't match. It wasn't that I was overlooking something. No, this oversight had happened in the design phase, long before I ever got involved.
It wasn't hard to figure out what had happened: When Ford had demanded the last-minute switch to their standard Kawasaki robot, the controls engineers had simply swapped the electrical drawings without thinking about the fact that the weld guns were not the same units that the standard had been designed around.