What you are wanting to do - it sounds like - is replace some industrial controls with a hobbyist-oriented board. So keep that in mind.
Can an Arduino do what you want to do? Certainly (for your number of connections, though, you might want to go with a Mega). Is the Arduino reliable enough for the task? Possibly - even probably. That will depend on a number of external factors, along with how it is set up in the environment. There are larger issues, though:
-
How to deal with the various industrial voltages - both for output, and input; usually with PLCs you are looking at both AC and DC, at a variety of voltage levels.
-
How to deal with any communications that need to happen between machines (5 volt TTL is probably not going to work well in an industrial environment).
-
What proof or certification will you have that a system based on the Arduino will work flawlessly under a given set of conditions, until the MTBF? Industrial controls go through an extensive testing and certification procedure, from what I understand...whereas a control system made with an Arduino - well, wouldn't.
-
Probably the biggest hump will be "liability" - should something go wrong with the system, who has liability? For instance, if someone is injured on the job by a machine being controlled by a standard PLC, and it was found that the machine failed due to a spurious signal from the PLC (or a mis-read of a sensor), where all other factors are accounted for as "working" - generally liability would fall to the manufacturer of the PLC. In the case of an Arduino replacement? Who knows - but probably whoever implemented the system. Then there's the larger question of "what does the company's liability insurance policy read?" - Does it specify that the policy only covers issues that arise due to faults with certified (by some named industrial certification group - like ISO, UL, or such) equipment? Would using an Arduino void the insurance claim in the event of a failure?
All that said - if you still plan to go down this route - there are some options and ways to make things as best as possible for industrial control.
First - check out Rugged Circuits - who sell "ruggedized" Arduinos and such. The owner also posts here in the forums on occasion. They have several products that will probably interest you.
Also check out Industrial Shields, who sell Arduinos designed for industrial PLC control. They might actually be your "go-to guys" for your application, as you will see once you browse around the site.
You might want to also google "arduino PLC", "industrial control arduino", and "rugged arduino" for more.
From a software standpoint, you can control things using C/C++ on the Arduino, or you might want to stay with ladder logic. Fortunately, there are options here for the Arduino - for example:
http://www.open-electronics.org/arduino-as-a-programmable-logic-controller-plc/
...and:
http://www.electronics-micros.com/software-hardware/plclib-arduino/
...and:
http://cpipero.github.io/ArduinoLadder/
I'm sure you can find more by googling "Arduino PLC library" and "Arduino PLC software". Another alternative might be to install Firmata on the Arduino, and control the Arduino using a PC running custom or proprietary software (provided it has a way to interact with the Arduino - whether through Firmata or some other library).
So, there are options available for the software side of things, too. Know that there are more than a few people working on the same problem(s) as you are finding - and have an interest in solving them and/or providing solutions (and help, I would imagine). You can even find a few threads here on the forums about industrial-level control and using the Arduino as a PLC (as you google, you should be able to find these threads, if you haven't already).
Still - the one "nut" that hasn't been cracked (due to the insane cost) is certification of the hardware for industrial control use. It isn't so much that the Arduino can't be made rugged and robust enough to handle the environment or the tasks placed on it - it's more that to get the needed certs (which again, might be needed for insurance purposes) - costs a fair amount of money for the testing (often destructive testing to boot). Look into what it costs (and the process) to get UL certified - that's the kind of gauntlet that would need to be run by companies trying to sell "Arduino PLCs" - it's also why industrial controls and PLCs that -are- certified cost so much; the costs cover the liability and testing, to an extent (there's also the niche product factor, too).
Good luck - and I hope I have helped. 