While I don't work with PLC programming specifically, MCUs work similarly. They nowadays implement a low voltage detect, LVD, aka brown-out detect, BOD as standard. Back in the days some >15 years ago you had to provide this protection yourself with external voltage supervising circuits.
Quite often there's a possibility to get an early warning when the supply voltage is dipping towards dangerous levels, but before the point where the part has to reset itself since it can no longer guarantee reliable execution. This is for "limp home" purposes such as shutting everything down safely etc.
Similarly, you should be able to see what was the cause of a reset - was it a power failure? This could be important for diagnostics and trouble-shooting purposes.
A PLC may in addition keep track of how much current a certain output function consumes and what it does during overcurrent would be hardware-specific. Quite likely the outputs have built-in overcurrent protection in the form of smart MOSFET drivers, but I wouldn't count on it.
I've actually seen plenty of automotive/mobile PLC which happily let its I/O circuitry burn to a crisp. Don't assume that the really expensive PLC was designed by competent people! I could name drop some big companies in mobile hydraulics who produce absolutely awful PLCs, that easily break in case of overcurrent/shorts, reverse voltage etc. Or go bananas when faced with voltage spikes, large current surges and other such highly expected EMI events in any industrial machinery. I've even reverse-engineered some of these PLCs and concluded that several of them were plain badly designed...
