Just a little curious

Hello all.


I was starting to get into a new project (that will probably never see the light of day because $$$) at the plant I work in and started wondering if I am going about this the best way. We have a machine setup that my boss would like to automate with a PLC and maybe an HMI.


So my question is how would you go about starting to take a machine that is currently controlled with switches, buttons, motor starters, relays, etc.. to something that is controlled by a PLC?


I personally started writing down all the current inputs and outputs that I can see would still be needed by looking through the electrical schematics we have. I figured what me next step would be is to figure out what other inputs or output may be needed, if anything. I this generally the best way or does anyone have some suggestions as to the steps to go about this process?


Your input is greatly appreciated.

Somer said:

Hello all.


I was starting to get into a new project (that will probably never see the light of day because $$$) at the plant I work in and started wondering if I am going about this the best way. We have a machine setup that my boss would like to automate with a PLC and maybe an HMI.


So my question is how would you go about starting to take a machine that is currently controlled with switches, buttons, motor starters, relays, etc.. to something that is controlled by a PLC?


I personally started writing down all the current inputs and outputs that I can see would still be needed by looking through the electrical schematics we have. I figured what me next step would be is to figure out what other inputs or output may be needed, if anything. I this generally the best way or does anyone have some suggestions as to the steps to go about this process?


Your input is greatly appreciated.

Click to expand...

Good first step.

Simplify would be another word to think of.

And, if it ever gets serious, The Naming Of Things is a big deal...better to argue at the start, than fix at the end.

You can search for a good article that's still floating around called The 10 Mistakes People Make When Purchasing Automation.

Don't get into the details too early in the process.

The first step would be figuring out the benefits of converting the controls to PLC-based. Can you improve performance by producing more per hour or by making the products more uniform or to tighter tolerances? Reduce waste? Make it more reliable? Reduce maintenance? Make it possible for one operator to tend multiple machines?

If the machine is currently in operation, spend some time watching it run. Ask the operators questions about how it works. Ask for his suggestions about how it could be made to work better or how to make it easier for them to make good product with it.

steve bailey said:

don't get into the details too early in the process.

The first step would be figuring out the benefits of converting the controls to plc-based. Can you improve performance by producing more per hour or by making the products more uniform or to tighter tolerances? Reduce waste? Make it more reliable? Reduce maintenance? Make it possible for one operator to tend multiple machines?

If the machine is currently in operation, spend some time watching it run. Ask the operators questions about how it works. Ask for his suggestions about how it could be made to work better or how to make it easier for them to make good product with it.

Click to expand...

+1

jdbrandt said:

And, if it ever gets serious, The Naming Of Things is a big deal...better to argue at the start, than fix at the end.

Click to expand...

I like you.

Somer said:

Hello all.


I was starting to get into a new project (that will probably never see the light of day because $$$) at the plant I work in and started wondering if I am going about this the best way. We have a machine setup that my boss would like to automate with a PLC and maybe an HMI.


So my question is how would you go about starting to take a machine that is currently controlled with switches, buttons, motor starters, relays, etc.. to something that is controlled by a PLC?


I personally started writing down all the current inputs and outputs that I can see would still be needed by looking through the electrical schematics we have. I figured what me next step would be is to figure out what other inputs or output may be needed, if anything. I this generally the best way or does anyone have some suggestions as to the steps to go about this process?


Your input is greatly appreciated.

Click to expand...

An I/O listing is a good start. Of course, an HMI would make most of the hardwired operator interface I/O (pushbuttons, lamps, gauges, etc.) irrelevant so far as your PLC's I/O capacity requirements.

Steve's suggestion to familiarize yourself with the process is also critical to a good outcome. All too often, I have witnessed machine operator's feedback in such upgrade situations being ignored by engineering staff, although a good operator can share a wealth of information about how a machine operates, since they live with it every day.

Besides, you'd rather have the operators on your side, than have them working against you when you're getting up and running after the conversion.

In today's climate, it's also a good idea to review the safety considerations for your particular process.

Well the "machine" in question is actually our chiller system. It has redundant circulating pumps that, in the 5 years I have worked here, never get used. The idea is to change the system so the pumps will be cycled so one pump runs for x hours then the next pump will start up for x hours to let the first pump "rest". This should in turn give a little more reliability for the system. We had an issue in the past where a pump went down due to a burned up motor and we couldn't start the redundant pump because it was seized in place.


In my mind it is a fairly "simple" thing to do. I was just seeing if anyone had some input on the process I had started going through to figure out what is needed and stuff.

ASF said:

I like you.

Click to expand...

Me too.

why would your boss not just ask you to switch the pump manually every week on a work order and not need to spend money

cjd1965 said:

why would your boss not just ask you to switch the pump manually every week on a work order and not need to spend money

Click to expand...

I know. That is the simple solution. I think he came up with this as a way to keep me busy and off his back about other things.



I do appreciate that it keeps my brain active instead of droning on with the mundane blown fuses and operators complaining that "this doesn't work the way I think it should."

jdbrandt said:

the naming of things is a big deal

Click to expand...

a++++++

If your drawings are accurate, you can recreate a like for like functioning of it by copying the schematic into ladder logic.

However, when presented with an opportunity like this you should fall back on the drawings for interlocks and safety features (which hopefully you'll increase) and specify before any code is written how a better machine would work when you put the PLC on it.

Also worth thinking about is the interface... there's a lot to be said about buttons and lamps if the system is particularly important. An HMI is great, but in some systems we do want something better than that.

I think posters here have given you many ideas, I will just re-iterate (and possibly add, however, I have not read all replies).
Get the existing control schematic and come up with a minimum I/O list, think about extra I/O for improvements, allow at least 10% for future expansion.
Think about what you don't need, i.e. if fitting an HMI, probably do not need as many buttons (I/O).
Come up with a requirement spec (what it needs to do), create a functional specification (how it will work). Think about the components (what spec does the company use i.e. PLC/HMI type and other components), if they have no spec try & come up with one, suggest use components/suppliers that you have a good relationship with and components that are readily available.
In the past I have done this many times, at one point I was given 3 projects to look at even though I was sure they would not go ahead, I did everything from design & producing documentation including writing & simulating the process. To my surprise the largest one went ahead, how surprised they were when I said I was ready for the mech/elec could go ahead as everything was ready, a bonus was it improved my understanding of the process.

If you change some activity on the machine that is being done at the moment only by user-interaction to being more or less automatic, then it is definitely a significant change that triggers that the risk assessment must be updated - or made from new.

Since you are in the US, as far as I know, in the US persons that design a control system for a machine can become personally liable if an accident occurs and it is revealed that the accident could have been prevented.

jdbrandt said:

And, if it ever gets serious, The Naming Of Things is a big deal...better to argue at the start, than fix at the end.

Click to expand...

You can say this one again!
This is an ever lasting battle between myself and the folks that work on and install my machines and process installs. If I have half a dozen photoeyes that are labeled PE1563 and PE1573 and so on, to match the line numbers in the drawings. It makes me almost feel like going psyco when I come to the site and find the devices and wiring labeled all the way through as PE1, PE2.... Then they do the same with motors. Make up stupid names when the line number was already designated as the identifier.... I cant beat this into some people. Its a big deal on a large process and usually will haunt the system and its techs for its entire lifespan once up and running.

Big rant yeah, this has been the theme of the month(s) here!