Process and Control Engineering

When it can't properly control the process. So far automated cars are not quite there (but getting close). On the other hand automated turbine control in power or gas plants is pretty reliable.

Process control is getting more and more automated, because the processes are becoming more complicated. An automated system can do process optimization, human operators don't do as good a job of that.

When it's safer, cheaper, quicker, easier, or more reliable (to name a few) to do it manually.

(When it's time for a bed-bath.

When it's time to tuck them up in bed and say 'night night ...')

The list goes on.

Do you have (a) particular scenario(s) in mind?

Im asking when Automation is not appropriate in any scenario so I can get a wide range of responses.

When was it not appropriate for you???

sustainergy,

The posts seem to be answering your question from the technological or cost accounting side, and I have read many good answers. Some answers from the social side may also be in order:

1. When the automation is harmful to the people or the structure of the society. (Consider that some respected social groups eschew many "modern" or "labor saving" products because they are seen to cause isolation from each other or from the natural environment.)

2. When the benefits of the automation are not available to the people who have lost work or jobs because of it.

3. When the effect of the automation is to reduce the workers to a status of human machines because of the speed or the impersonal nature of the work remaining.

4. When automation is implemented without carefully considering how people will interact with the automated process. (Consider a prompt to enter a number but the program does not exclude non-numeric input, or perhaps the program does not help the person to give the necessary answer or waits without any human-understood reason.)

--JMM

My reply - better late than never!

Automation is not appropriate for anything I do (PC design work excepted), as the deliverables I am involved in producing are never in batches of more than 5, and require a lot of wiring of discrete components (relays, connectors, PLCs etc.).

Much of the stuff I produce, however, is used to automate processes - so it is appropriate for my customers. It frequently involves automating a system for measurement involving radiation - either calibrating instruments by irradiating them, or directly counting the radiation emitted from samples. In either case, the measurement times are often lengthy (5 to 10 hours per measurement not being unusual), and the measuring equipment is highly capital intensive (costs a lot!). The customers find it appropriate to automate because they can increase their throughput without the additional expense of either employing people out-of-hours (during the night and at weekends, when the rest of the establishment is unmanned), or duplicating the sometimes very expensive equipment.

Other things I've been involved with requiring automation are remote handling systems, where the prime objective has been to reduce personal exposure to some hazard.

Beyond these examples, I can't comment from personal experience about automation in manufacture.

AUTOMATION is all about minimising total per-unit cost. Total production cost falls into two areas: non-recurrent engineering (NRE) to design, build and deploy the production system; and variable cost, which captures labour, materials and any other expenses that increase as production volume goes up. Automated systems trade increased NRE for decreased variable cost.

Theoretically, any individual manufacturing task or step can be automated. With todays technological level, we are very close to changing the world "theoretically" to "practically." There are few tasks that can't be automated.

Cost and pricing are big issues to consider. Design your process along both lines and see where you can cut on cost. Automation is dictated by your process; and by the way things are going you can manufacture anything anywhere by combination of both. I have worked for some of the finest companies in the world and they seem to have done greatly by using one or the other and sometimes both. Study and revise your process; see whe you can make significant changes and reduce cost. Just remember one thing. Quality is built into a device. This can not be substituted for under any of the circumstances.

My first thought was: When it's inappropriate! Sorry, sometimes I just can't stop myself!

1. When it reduces whole populations/societies to practicing the phrase: "Do you want fries with that?"

2. When it isn't economically feasable, just the wish of either an asocial manager, or a socially inept "techie".

3. When it is used as a way to more easily bypass safety. That needs an explanation that is a subject unto itself, but there are some basic principles. Historically, equipment and personnel safety interlocks have been "hardwired". Some still are, for example: The lanyards on conveyor systems. But over the last 30+ years I have seen more and more safety systems moved into the automation circle of influence. Unfortunately, when that happens, it is easier to change, and an unsafe/clueless supervisor has to convince only one person (the programmer) in order to get a safety interlock bypassed, to make his production goals - safety be damned. That can be done in the hardwired world too, but the "jumper"is usually a dead give-away, and there are usually more people (co-workers) that are made aware of it. The tracking of software changes is getting better, but "it ain't there yet".

Your title is process and control.

So it is not appropriate when the process that, is hard for automating or is too expensive or not safty if it was automated.

More further, when we want to do any project we have to evaluate the effect of it e.g. profit, productivity, safety,....

Good luck and success with your automation.

Over shot or under modify in order that adjustor is very slow to get a right condition.

sometimes a good parameter needs experience to set a suitable parameter.

Automation is not appropiate when the unit cost increases without affecting Quality and safety.

I work with compression molding machines, and operators need to do qc before cnc sanding. It is easy to fix minor machine problems when an operator is right there. Operators are better and cheaper at detecting smoldering material and easily adapt to different parts on a line (i.e. no reprogramming or debugging).

1) When there is no system which can create autometion for complex job. 2) When reliablity of Automation is not guaranteed and there is possiblity of failure of automation resulting in big losses.3) When Automation can create problems of human replacement and unemployment in labour surplus countries. 4) When its cost is very high and not compareble with exisisting systems even in future. 5) When it can not match up with other production activites and causes imbalance in production lines.

In whatever circumstances...it is the COST... this cost inclusive of quality, safety,deliveries.. even human life...if the cost of losing life is still cheaper then automation is not appropriate. Because automation is not meant for quality...you may produce 90% junk but the 10% still bringing in more then investment on automation ROI, then automation is not appropriate

Automation is not appropiate if is meant to enable operators to sleep in the control room or to make the process idiot proof. It is also not appropiate if the cost for installing the system is higher than the cost of system failure.

It all depends on whether or not you are building a new processing plant or upgrading an existing process. I would always automate a new processing plant wherever possible. An upgrade will require some value engineering whereby one looks at what the value is when implementing automation.

I also think that one must always look at the root cause of a problem and address that first rather than simply automating something down the line.

I work in the Tulsa Oklahoma area, and just as an example of when automation goes wrong, I am relaying what happened in this area.

A worker at a large well known uniform company was doing his daily job of making sure the clothing going up the automated conveyor and into the huge washer dryer.

A jam-up occurred at the top of the conveyor. Somehow the safety interlock for this process had apparently been bypassed or was not functioning. The worker taking the situation into his own hands went up to where the jam-up had occurred and tried to kick the laundry into the machine. The conveyor knocked him into the machine where he was subjected to the high temperature wash cycle and the beating of being tossed around in the machine as well as the drying cycle where he was subjected to around 475 degree temperatures for a period of 39 to 45 minutes.

No he did not survive. His co worker had gone to the restroom when this happened and had no idea what happened to the man and it was not discovered until it was way to late. (after the cycle had ended.)

So i put it to you was automation a necessary thing here? I don't think his family thinks is was.

Don't get me wrong I feel the need for automation to get the cost of everyday needs down is a good thing. I also feel like it is necessary in a lot of processes. We use it every day her where I work. I think it needs better monitoring where safety is concerned and not less. Especially when peoples lives could be at stake.

End of my 2 cents.

This is a tragic event, but the root cause lies in the operating procedures, NOT the automation. I see two significant issues - "safety interlock ... had apparently been bypassed or was not functioning" and the worker intentionally putting himself in an unsafe position. Too often, people tend to put the blame on automation because it's what they understand least, and it allows them to avoid the uncomfortable situation of having to reprimand a person. Imagine in this case if there were no automation - a worker would have to be manually loading the machine, thus being in the same unsafe position that resulted in the death all the time. With automation, he can stay safely away from danger. It is not the fault of the automation that the safety interlock was not engaged, or that the worker committed an unsafe act, either in defiance of or due to lack of safety procedures.

Although I agree with you on this being more the fault of the person and not the system. The issue of safety and safety training is the root of the cause. I believe after having talked with a couple of the people that worked for that facility, that the system started up automatically because there was not a safety interlock device in the area he was trying to clear.

I may not have said that in my earlier post. The worker I believe thought he was safe in what he was doing when the machine started up. But who is to know for sure as no one knows what the man was thinking when he did what he did.

I do not fully blame the system and I am always appreciative of automation in every aspect of production, except where the final product needs to be inspected by the human eye. No machine cann fully replace the human individual but many can be made to minimize the number of people it can take to operate them.

So please don't take it that I am against automation. I just want to make sure when they are designed that they are made as safe as possible to prevent possible injury or death. I know they can't be made idiot proof because there is no such thing.

Thanks for pointing this out DATA. Many of the these comments for when its inappropriate are based on errors by people and have nothing to do with when actually automation is inappropriate.

Safety Interlocks should 'NEVER' be processed through a PLC/Computer where it is left for a programmer to make a decision. They should always be 'hardwired' to kill all forms of energy and where gravity comes into play, pin/secure it.

Quality issues arrive, not due to automation, but a lack of fortsight in designing the system properly. That is why there are FMEAs in both the design & process stages. They are suppose to be living documents that tells us what possibly can go wrong when the tools (manual/auto) we use come in contact with the product we are trying to produce. It helps us in focusing in on the higher risk areas that might allow bad product to get past our detection systems and out to a customer.

Automation is 'inappropriate', as has been mentioned multiple times previously, when it not cost-effective to do so. Also, has many have said but may not have properly explained, it is not appropriate when the manpower capability is not there to properly implement it. Even if you hire outside help, what expertise do they have in your product? Very little, so without the knowledge in-house to provide quality input to the outside automation house, you're in the same boat.

One of the biggest problem areas are when we try to use it in quality inspection areas. Why? Not because the automation can't do what we tell it do, but because it becomes very difficult on people programming to pre-determine every little issue that may arise in a process. A properly trained person can adept much easier to unique changes in the process then can the vision system.

when it's only done once . . .

I remember a story about two old guys watching a steam shovel dig a hole for a foundation. One old guy lamented that "one hundred men with shovels could have done the same job".

The other old guy responded "then why not a thousand men with teaspoons".

Automation is simply another tool to increase productivity. Just because this tool has wires does not make it less appropriate.

Most of what I design/build is at a maximum run of 20 quantity. If I were to spend the money and time to automate everything, I would have some fancy equipment, and no money, parts, or space to manufacture products. Now if I were manufacturing 1000 doohickies a month, then automation might look more interesting.

Now I do have parts which are common to many various antennas which I manufacture. These I fabricate in quantities of (typically) 100 at a time. In doing these one at a time, I would spend more time setting up the drill press, saw, etc. than I would spend making the part. This might be considered automation although it does not use computer aided manufacturing... well... in another way it does. Let me explain.

I am currently setting up to make "shorting bars". These are rectangular dice 1/4" by 1/2" by 1.2". I start out with aluminum bar 1/4" by 1/2" by 12 feet long. I cut the bar in half just for ease of handling. Each shorting bar requires a 1/4" hole and a 1/8" hole drilled through it. I have layed out a template on the computer which has the locations for each hole. I print out, cut out and glue the template to the aluminum bar. I will then set up the drill press to drill the 1/4" holes, start at one end of the aluminum bar and drill to my hearts content. I then put in the 1/8" drill bit and drill all the 1/8" holes. Now I take the aluminum bar over to the saw and cut the individual dice. The saw cut locations are also on the original template. There are other operations involve in addition to those described, but this has the basics, and in the end I have 100 shorting bars, and a bit of scrap for me to play with.

It isn't fancy automation, but it works for the small quantities I deal with.

Bill

Hi Sciesis2,

Ever thought of linking the pressdrill and the cutter saw together? With the right distance/alignment, you could do both operations at once. Just a thought....

Rick

PS Happy Holidays to you and yours. Stay safe.

Rick

I would love to do something like that, but it is a matter of quantity and room. If I were making 1000 per month, I would love to set up a situation where I could feed 12 foot aluminum bar at one end and get finished shorting bars out the other. In fact it would be fun to take on such a challenge. But the quantities I deal with don't warrant the time and expense to set up such a system... Maybe when my ship comes in.

Bill

Automation is at anytime, anywhere.

We cannt keep off it. we need it always.

just imagine, how shall we live if there is no automation?