The asians got this one wrong.  As I stated earlier in this post, safety requires forward thinking.  If a wire gets cut- we should predict the state.  No sane person won't understand that a wire that got cut will very likely go to ground or be cut clean......  It could short to another hot wire but the odds are stacked way against that.

So, when the asian drive folks opt for grounding the input to make it run, they clearly haven't progressed through the lawsuits.

Most PLC's can do the bank of X number of inputs or outputs.  For example, Inputs 0-3 may share the same "common" (not to be confused with DC Com).  If you connect the DC Common to this terminal, this provides a path to ground or specifically DC Common.  So if 24 VDC shows up on any one of 0-3, it will allow current to flow to the "common" terminal because you connected DC Common or ground to that terminal. Howerver, if you hook up 24vdc to the common terminal.  In that case, when you apply 24 vdc to one of those inputs (0-3), no current will flow because 24vdc is on both sides of the circuit.  The Input will not be activated.  In the same setup (24 on the common),  If you ground or connect dcComm to one of the inputs, current will flow "backwards" (thats highly debatable) and the input will register.

It is very useful to have banks of inputs that are selectable for sinking or sourcing.  Case in point is the Automation direct drive I have.  It will only provide a sinking output (meaning it provides a DC Common when ON).  The only way to use this directly is connect 24vdc to my "common" for some input and let the drive provide a ground or DC Common.  This will activate the input.  This is why CSLabs has all those dedicated connections for each input.  If, Like PLC manufactures generally do, they banked 4 at a time, I would waste 4 inputs if only one of my inputs required a sourcing input (meaning current needs to flow out of the input- to the drive------highly debatable again).