Back in the 90s when I had just graduated with my mechanical engineering degree, I took a position with a small, privately-owned company building electric buses and shuttles. The company's shuttles were used in many locations such as Santa Barbara, Santa Monica, Burbank, and more. They made some good decisions in that they kept things fairly simple, and they managed a 12 to 14-hour run time on a single battery pack by continuously decreasing the weight of the bus while increasing the capacity of the battery.
While working there, I suddenly found myself responsible for seven half-finished bus projects all at once, and due to recent employee turnover, the projects were riddled with problems. There was no one I could turn to for any tribal knowledge, since all the former mechanical engineers were no longer around. I ran into one particularly nasty problem with the transmissions. They were failing with unusual regularity on our 26 food transit shuttles. There were dozens of these vehicles in service, so I needed to fix the problem quickly.
The drivetrain of the bus was fairly simple. There was an electric motor coupled to a chain transmission, such as the Morse/Emerson silent chain style. This was a single-gear reduction, flat-chain gearbox. The output of the transmission was coupled to a driveshaft, which in turn was coupled to a Dana 80 axle. The combination of the transmission and the differential on the axle provided the necessary gear reduction for the high-revving electric motor.
The first problem I came across was that technicians in the field were replacing the recommend gear oil in the transmissions with a thicker oil in order to "help" avoid failures. Thicker is better, right? The problem with this approach was that the oil they were using was not preferred for this style chain. The oil needed to have specific friction characteristics to lubricate the small pins that traverse the chain links. The heaver oil could not get into the crevices to properly lubricate the chain. Even if it could get in, it did not have the proper friction modifiers to lubricate the chain to the manufacturer's recommendations.
This improvised "fix" only served to accelerate the failures. It also increased the noise. Even though it is counterintuitive that thicker oil would make a transmission louder, the much-thinner recommended synthetic oil greatly reduced friction and allowed for a much quieter transfer of power.
Now that the oil dilemma was resolved, the transmissions were still failing. The original problem had yet to be discovered, and I still had some work on my hands to figure out why. As I observed a new shuttle being built, I noticed something interesting. The Dana axle was being installed pinion-down. Having spent time researching the axle with Dana, I knew it was a pinion-up design. The fabricators and assemblers in the shop were used to seeing axles in the pinion-down position and just assumed that was how it was to be installed. Adding to the problem, the production documentation was, unfortunately, not detailed enough to have helped avoid this simple mistake.