Control Charts + Ishikawa

[Jim Wynne]

Welcome to the Cove.
<snip>

So my question is: How could I know both limits (upper and lower - I am asking about ALL parameters) in case if is NOT determinated anywhere (on the technical drawing)? More than 50% of machines don't even have the screens to check and even if they had them (screens) this wouldn't help me because they would show current parameters and not maximum and minimum allowed.

It seems to me that the first thing you need to do is classify the defects, understand the frequency or likelihood of each one (and associated costs, if practical). Given that information, you then need to identify the process characteristics that contribute to those defects.

The extent to which process characteristics may be controlled is the object of control charts, and control limits (not specification limits) are a product of the data. The type of control chart to use is dependent on the nature of the product and process.

Besides my main question, I also have two subquestions if I could ask...:

1. Is 50 bolts per day per one machine good daily amount? If yes: what should be relation between good and bad parts? I was thinking about 35 good and 15 bad. I think its a good idea. May I please for your comment, what do you think?

I'm not sure what you're driving at with regard to the relationship between good and bad parts. I think what you need to do is understand the ratio and determine what's causing the defects. The number of defectives considered acceptable is largely a matter of economics, and the potential need for sorting, and how your customers tolerate receiving defective parts.

2. I work daily from 7 AM to 3 PM which is classical working time for almost everyone. One of the sorting criterias will be based on worker (first + last name) who is responsible for machine. We don't have many workers so this should be a problem BUT another group of them start working from 2 PM to 10 PM. The only exception is the section for thermal processes who also work the night job (until 6 AM). Issue is obvious - I cannot be here 24/7. Sorting the cards based on specific worker should also include sorting based on time when they work. Am I absolutelly required to still sort based on work? Note: I have in purpose to choose the bolts for analyzes from different locations of the same box which says a lot. I ask this because workers' productivity and exactness is not always the same in the different working times (regarding exactly the same person).

Human influence is one of the process characteristics you'll need to understand. Sometimes it's best to start from a broad view and narrow it down as necessary. For example, it would probably be best to determine, if possible, if defects are most likely to occur during a specific shift, rather than trying to narrow it down to a specific person or machine from the outset.

As a former boss of mine was fond of saying, "If this were easy, anyone could do it." It's not easy, and there are many variables and vagaries that must be understood if you're going to avoid making things worse.

If anyone knows any good book / online tutorial on this topic, I would please for suggestion.

I highly recommend Donald Wheeler's Understanding Variation as a starting point. You might also find the NIST/Sematech e-Handbook of Statistical Methods helpful. Note that the latter link is to a U.S. government site, and may be blocked to some visitors outside the U.S..

 

[Your Soap]

Thank you for your replies.

bobdoering all our machines except Essebi SB 508L are very old. We also have two Dimac MCV machines but those two don't count because they are just for automated controlling larger series of bolts so we get them separated as good and bad ones. However there is no way that I could do my diploma project with a help of Dimac machine because they just separate bolts and thats all. You mentoined the understanding of processes; after the bolts, without screws are made, there can be several options whats next. Usually next process is rolling unless the customer requires no screws on bolts. Also milling can be done for some special requirements of the bolts' top. Always is cleaning the bolts done. This is done on some huge machines, also very old, but I don't remember the exact model of them. I can check and report later. Thermal processes are very often done as well.
So yes; threads (screws on bolts) are created separately. In other, separated, location of the company. Both locations are near each other. Of course because of logistic reasons. You suggested me total variation equation. If I understand correctly, this should be done as first step before doing anything else such as taking out for analyzes at the first day the first group of 50 (I chose this number on my own, the amount might be wrong also) bolts? Its no problem to analyze dimensions on the technical drawings but I am very worried, as I asked in my "main question", about the parameters that are NOT on the drawings? Almost nothing is near machines. I don't want to bother the workers too much. I need to control them and their work. Also on the documentation papers are not mentoined those parameters such as force and, as total variance equation shows, temperature and many more.


Stijloor yes I apologize because of that. My English language isn't the first one, so called mother-tongue, so might be a bit difficult to type everything 100% correctly but hopefully I am understandable enough. Throw out of 2% is the value which can be dynamic and it depends on many parameters. When it comes to sorting the control cards, I am trying to include all of the parameters. Thats why I asked my second subquestion. I was typing the idea that I got how to, very well, cover two different parameters (exact machine and worker) with one section only - choosing only the bad parts from the machines with as much as possible throw out (bad parts). This way I can include together all the machines because I will analyze the problematic ones and at the same time I can analyze the worker. It might be issue on his side why so much throw out. If huge amount of throw out will be repeated by same worker but different machine then the problem might (or might not) be with worker and not with machine. There are also several possible issues what could be wrong on worker's side.


Bill Pflanz yes I will try to find this book or some other tutorials. Its not hard to draw Pareto diagrams, neither control cards but, since I don't know the limits of processes I cannot begin with this yet. I guess the author of the book (Wheeler) doesn't need to be the same. In the ideal situation I might find some online tutorial that ''walks'' the user step by step towards the final result: what / who causes the throw out. At least what I would like to start as soon as possible without doing bigger researches yet is drawing the control cards because it will take a while - approximately two months for good results.


Jim Wynne I didn't understand your reply what you meant? Gathered defects should be all parameters that depends on workers' job - you called them process characteristics. Am I wrong? For my opinion there is no way to determinate what is causing the throw out without doing some graphical analyzes to see what is running outside the limits. The expression "control charts" is what I meant with "control cards", we are just calling it like that. After I already know everything what should I control (so far I know only dimensions with tolerances on the drawings, hopefully the book will help to gather more characteristics what to control) then I will be able to start with the those charts/cards. But...

and control limits (not specification limits) are a product of the data.

Did you mean I should choose the upper and lower limits instead of trying to search inbetween which two values can specific machine work with giving out the best pieces? But even if so, my choice still need to depend on something, I cannot just guess the limits.

 

[bobdoering]

Thank you for your replies.

bobdoering all our machines except Essebi SB 508L are very old..

I believe this is a cold heading machine. He is describing making bolt blanks with this machine.

For any control, it is not possible to control every possible characteristic - either on the part or process parameters. So, you need to pick out the most important ones. For cold headers, fill (OAL typically, because it is the longest draw) and die match (head to body runout - ensuring top and bottom of die are properly aligned) are important. Most other dimensions are controlled by the die cavities and have no adjustment.

So, what are the variations that contribute to the OAL length (die fill) total variation? Length of wire cut, cleanliness of cut, wire chemistry, wire diameter variation, wire hardness, die force and speed, temperature (any preheat?), temper, etc. Ask the set-up guys who know the machine if there are any other causes of variation or adjustments. Sort out the ones you can adjust (such as wire length), those you can not (hardness) and those you can maintain constant (die cavity dimensions).

My bet is you can use SPC (control charts or "cards") on the OAL length to assure the process is running OK. Best bet, I-MR chart. I would not bother with Xbar-R chart, as the variation between 5 parts will be insignificant. What should the specification be? Once the machine is running well (after warm-up, as warm-up is a "special cause"), do a capability study. How long is a typical run? 5000 pcs, for example? Take a sample every 50 pcs. You want about 100 data points spread out evenly across the run.

Usually next process is rolling unless the customer requires no screws on bolts. Also milling can be done for some special requirements of the bolts' top. Always is cleaning the bolts done. This is done on some huge machines, also very old, but I don't remember the exact model of them. I can check and report later. Thermal processes are very often done as well.
So yes; threads (screws on bolts) are created separately. In other, separated, location of the company. Both locations are near each other. Of course because of logistic reasons.

Threads, how painful. For roll threading there are 2 key issues - set-up and assuring the entire part is threaded (as in inserted all the way into the machine). What thread gaging do you have handy? How do they check their threads now? That will make huge impact on determining the data from roll threading!

You suggested me total variation equation. If I understand correctly, this should be done as first step before doing anything else....

Yes! You need to understand the variation, what you can adjust, what you can keep constant and what you can not control ahead of time before diving into the machine. Otherwise, when you see variation coming out of the machine, what are you going to do about it? Sure, as you get more data or understanding of the process, you may revise and correct your total variation equation - but you need to think about the process before diving into it!

I asked in my "main question", about the parameters that are NOT on the drawings?

What you don't have specifications from you need to develop from capability analysis.

Almost nothing is near machines. I don't want to bother the workers too much. I need to control them and their work. Also on the documentation papers are not mentioned those parameters such as force and, as total variance equation shows, temperature and many more.

Grab your samples, and take them away from the machine - no problem. Just be sure you identify them in time order - not just a big pile. Make a rack to keep them in order - it will help a lot. You must keep your raw data in time order for complete analysis! Jot down any process parameters - especially the ones the set-up people or operators can adjust.

One last word or warning - you will not get every possible variation, no matter how hard you try. Even the location of the cooling lines can affect the results! So, be prepared to do the bust you can, but do not expect a perfect outcome!

 

[Your Soap]

On Essebi SB 508L the top of the bolt won't be aligned exactly to the center because the bolts there are made on purpose that the furrow at the top is more close to the edge. But it would take hours to search for bad pieces on Essebi so I think I must find another way. What you meant with ''fill (OAL)''?

Length of wire cut depends how much (according to weight) is needed for specific amount of pieces so this cannot be controlled regarding the throw out. When it comes to wire for bolts, I will control the material (with this is included wire chemistry) and provider. Why should I control the diameter of wire? This depends on the types of bolt - to be more exact it depends on diameter of ''rod'' (rod = part of the bolt which is under the top). Wouldn't be controlling the diameter of a wire included in sorting the control charts based on similarity ob bolts - according to the different types?

I don't think I could adjust the hardness of the wire. We get what the provider sends. The only thing I could to is suggest to the owner of our company to fire the "bad" provider - provider with not as good as we require wires.

What you meant with ''die force''? I have been looking near machines but there is nothing I can do with force. Temperature? Where?

I think the first few pieces after the warm up should not be included, not in the amount of created bolts (workers' care) and not in my analysis (my care). Workers first need to make sure that the first pieces in the range are running well but this doesn't mean, after that, they can go away from the machine. I should also take care of the amount of first pieces because as few as possible should be made for testing purposes at the beginning only so the throw out (lose of material) is as small as possible.
I am not sure what you meant with ''OAL length''. Probably dimensions that are defined on the technical drawings. If so, I thought to control everything with control charts and not only those dimensions and that's why I was and still am worrying for the upper and lower limits. I got myself a book Understanding Variation which suggested me the member Jim Wynne. I think I-MR chart is well explained there. So I can learn from there.

Typical run depends on the amount of bolts we have to do in specific range for specific customer. Usually for one palette. I would say around 250 000 bolts for one palette but this varies in the dimensions. It can be much more.

Threads, how painful. For roll threading there are 2 key issues - set-up and assuring the entire part is threaded (as in inserted all the way into the machine). What thread gaging do you have handy? How do they check their threads now? That will make huge impact on determining the data from roll threading!

The biggest issue for rolling is that they throw ALL the bad pieces in one bigger box. So there is no way to know which parts belongs to which range (for which palette) and to which types. Measuring every single one, even if I have guarantee that the part is counted as bad one, is a time wasting. So I still need to get the idea how will I deal with that.
I have problems with understanding all the English technical expressions. You mentioned ''thread gaging''. I presume you meant with this the thread controlling to ensure all the requirements are met. They are doing intermediate control with the measuring instruments. Threads are sometimes also checked with microscope (regarding angles).

Yes! You need to understand the variation, what you can adjust, what you can keep constant and what you can not control ahead of time before diving into the machine. Otherwise, when you see variation coming out of the machine, what are you going to do about it? Sure, as you get more data or understanding of the process, you may revise and correct your total variation equation - but you need to think about the process before diving into it!

Even when it comes to the dimensions on the technical drawings (the most ''easy'' subjects to control with control charts), there is not much I can adjust. Process must be always made with respect to the drawings. What can I adjust, what not and what I can keep constant isn't so easy to understand for me. I don't have so much experiences. What to do when I see the variation comes out of the machine depends what the graph will show (specific part of control chart). If it will linear increasing then the issue might be with this worker - another proof of this could be if the linear graph will start immediately decreasing. This means, to the machine, came another, ''better'' worker. If the issue isn't because of worker then something on the machine needs to be adjusted. At the same time I will check other control charts - it might be a problem with wire material.

You also mentioned capability analysis to know which other parameters should I control. This cannot be done with control charts and also ''capability analysis'' doesn't mean I will get the limits, does it? If I understood you correctly, with capability analysis I will get a result (list of specifications) what am I able to adjust so I can control this. I have done some research on google regarding keywords ''capability analysis tutorial''. I found some site explaining this could be made with Pareto diagrams but I, personally, strongly disagree with this. I think there is no way I could get specifications, through capability analysis, with Pareto. Another tutorial was explaining I should include, in this analysis, Cp and Cpk. I still think there is no way to get specifications from this. I know few of specifications that I already mentioned but need to gather all. Cp and Cpk are used when we know the process is not in control, which is totally different than analysis the capability for the processes. Tutorial such as

www.morris.umn.edu/~sungurea/statlets/usermanual/sect7_2.htm

(I am not able to post the correct link because of the forum rules)

proves that I am correct. For what it shows, I need to know with which specification I am dealing with - strength is an example on this tutorial. We probably both agree that, with capability analysis I should find specifications (all of them and not just those few that I know already) what to control.

You mentioned ''time order''. I thought I would take the pieces (bad and good one) one per day, so daily but the good ones should be taken from the same box from different ''locations'' (different depth of the box). With this I would cover different times. Due to amount of machines its not possible to do what the drawing says - controlling few parts once per hours. I will control only machines with as much ( ! ) as possible bad parts in the box. Another reason for this is that the amount of pieces in the box with throw out doesn't increase so fast (per one hour). I would be lucky if it increases daily.

 

[Jim Wynne]

On Essebi SB 508L the top of the bolt won't be aligned exactly to the center because the bolts there are made on purpose that the furrow at the top is more close to the edge. But it would take hours to search for bad pieces on Essebi so I think I must find another way. What you meant with ''fill (OAL)''?

Length of wire cut depends how much (according to weight) is needed for specific amount of pieces so this cannot be controlled regarding the throw out. When it comes to wire for bolts, I will control the material (with this is included wire chemistry) and provider. Why should I control the diameter of wire? This depends on the types of bolt - to be more exact it depends on diameter of ''rod'' (rod = part of the bolt which is under the top). Wouldn't be controlling the diameter of a wire included in sorting the control charts based on similarity ob bolts - according to the different types?

I don't think I could adjust the hardness of the wire. We get what the provider sends. The only thing I could to is suggest to the owner of our company to fire the "bad" provider - provider with not as good as we require wires.

What you meant with ''die force''? I have been looking near machines but there is nothing I can do with force. Temperature? Where?

I think the first few pieces after the warm up should not be included, not in the amount of created bolts (workers' care) and not in my analysis (my care). Workers first need to make sure that the first pieces in the range are running well but this doesn't mean, after that, they can go away from the machine. I should also take care of the amount of first pieces because as few as possible should be made for testing purposes at the beginning only so the throw out (lose of material) is as small as possible.
I am not sure what you meant with ''OAL length''. Probably dimensions that are defined on the technical drawings. If so, I thought to control everything with control charts and not only those dimensions and that's why I was and still am worrying for the upper and lower limits. I got myself a book Understanding Variation which suggested me the member Jim Wynne. I think I-MR chart is well explained there. So I can learn from there.

Typical run depends on the amount of bolts we have to do in specific range for specific customer. Usually for one palette. I would say around 250 000 bolts for one palette but this varies in the dimensions. It can be much more.

The biggest issue for rolling is that they throw ALL the bad pieces in one bigger box. So there is no way to know which parts belongs to which range (for which palette) and to which types. Measuring every single one, even if I have guarantee that the part is counted as bad one, is a time wasting. So I still need to get the idea how will I deal with that.
I have problems with understanding all the English technical expressions. You mentioned ''thread gaging''. I presume you meant with this the thread controlling to ensure all the requirements are met. They are doing intermediate control with the measuring instruments. Threads are sometimes also checked with microscope (regarding angles).

Even when it comes to the dimensions on the technical drawings (the most ''easy'' subjects to control with control charts), there is not much I can adjust. Process must be always made with respect to the drawings. What can I adjust, what not and what I can keep constant isn't so easy to understand for me. I don't have so much experiences. What to do when I see the variation comes out of the machine depends what the graph will show (specific part of control chart). If it will linear increasing then the issue might be with this worker - another proof of this could be if the linear graph will start immediately decreasing. This means, to the machine, came another, ''better'' worker. If the issue isn't because of worker then something on the machine needs to be adjusted. At the same time I will check other control charts - it might be a problem with wire material.

You also mentioned capability analysis to know which other parameters should I control. This cannot be done with control charts and also ''capability analysis'' doesn't mean I will get the limits, does it? If I understood you correctly, with capability analysis I will get a result (list of specifications) what am I able to adjust so I can control this. I have done some research on google regarding keywords ''capability analysis tutorial''. I found some site explaining this could be made with Pareto diagrams but I, personally, strongly disagree with this. I think there is no way I could get specifications, through capability analysis, with Pareto. Another tutorial was explaining I should include, in this analysis, Cp and Cpk. I still think there is no way to get specifications from this. I know few of specifications that I already mentioned but need to gather all. Cp and Cpk are used when we know the process is not in control, which is totally different than analysis the capability for the processes. Tutorial such as

www.morris.umn.edu/~sungurea/statlets/usermanual/sect7_2.htm

(I am not able to post the correct link because of the forum rules)

proves that I am correct. For what it shows, I need to know with which specification I am dealing with - strength is an example on this tutorial. We probably both agree that, with capability analysis I should find specifications (all of them and not just those few that I know already) what to control.

You mentioned ''time order''. I thought I would take the pieces (bad and good one) one per day, so daily but the good ones should be taken from the same box from different ''locations'' (different depth of the box). With this I would cover different times. Due to amount of machines its not possible to do what the drawing says - controlling few parts once per hours. I will control only machines with as much ( ! ) as possible bad parts in the box. Another reason for this is that the amount of pieces in the box with throw out doesn't increase so fast (per one hour). I would be lucky if it increases daily.

You're faced with what seems to be a wildly out-of-control system, and I think you're trying to do too many things at once. Until you understand the defects and the likely causes of them, you probably won't be able to do anything except make things worse than they already are.

As I suggested in an earlier post, you need to identify and classify the defects. No statistical analysis will help you until you do that. You say above that you lack the experience to be able to identify the process characteristics that contribute to the creation of defects. This means you'll need help--you won't be able to do this alone. We can help with that, but you really need someone in your plant who understands how things work there. This is because your processes may have been designed to create defects (inadvertently) and you probably won't be able to change what needs to be changed on your own.

 

[Your Soap]

So if anyone could help me regarding my three questions in the first post, written on 14.4., I would really please so. Its a huge project and I have to begin with gathering the datas as soon as possible because I have a limited time for a university's practical work. Thank you in advance.

 

[Stijloor]

So if anyone could help me regarding my three questions in the first post, written on 14.4., I would really please so. Its a huge project and I have to begin with gathering the datas as soon as possible because I have a limited time for a university's practical work. Thank you in advance.

How about the various responses provided in this thread? Did they help?

Stijloor.

 

[Your Soap]

I am trying the book Understanding Variation. It has a large topic regarding control charts. Also I am searching for InfinityQS software which does a lot what I need. I haven't mentoined it yet. Have you heard about it? So far what I used to read about this software I got opinion it might be worth to try.
My 3 questions that I have had in my first post remains unanswered so I am still waiting if I could please anyone for assist regarding them.

 

[Jim Wynne]

So if anyone could help me regarding my three questions in the first post, written on 14.4., I would really please so. Its a huge project and I have to begin with gathering the datas as soon as possible because I have a limited time for a university's practical work. Thank you in advance.

You might not have gotten the direct answers you were hoping for because your questions indicate an inadvisable approach to the problems. Nonetheless, I'll give it a go:

The main question what I want to know is the following:

In our manufactury of bolts we have around 70 machines. They are all very old so there is no way to find (i tried already) the answer on google for each machine. I have in purpose to analyze the bad bolts in the little boxes near the few machines (not sure exact number yet) that have as much as possible (which is very bad but good for analyzing)
throw out. Daily I will take 50 bolts per ONE (!) box (which means per one chosen machine). Of course I will also take good pieces and not only bad. Both, good and bad, pieces I will take from different locations (different depth of box) of the same small box. The reason for this is simple: I will analyze different time intervals. The technical drawings
require once per hour but there is no way I could do this - too many machines and I work just 8 hours per day. So I will analyze daily result. For two months.

Depending on results, I will draw control cards (I could do it with Pareto diagrams too). Lower limit and upper limit on the control cards is the same as lower and upper tolerance on the technical drawing BUT majority issue is the following:
I must sort all control cards based on many criterias and not only based on important parameters on the technical drawings (controlling the dimensions). Now where is the point? I have absolutelly no idea how should I determinate the upper and lower limit in cards for criterias such as force on the machine and everything else what the worker on the machine choose before starting the work. Noone will be willing write everything down for me two months, so I won't even ask them. There are many other parameters (not only force) such as space between the path where botls are coming from etc, where the limits are never determinated anywhere. We don't have manuals for every single machine and I haven't found values (maximum and minimum allowed) online.

So my question is: How could I know both limits (upper and lower - I am asking about ALL parameters) in case if is NOT determinated anywhere (on the technical drawing)? More than 50% of machines don't even have the screens to check and even if they had them (screens) this wouldn't help me because they would show current parameters and not maximum and minimum allowed.

It's not clear whether you're asking for specification limits (where none presently exist) for the product or for limits on machine settings, or both. In general, product specification limits are derived from end-use requirements--fit and function. If you have no specifications for a given characteristic, you shouldn't be trying to make one up without understanding all of the things that contribute to fit and function for the characteristic in question.

The settings for process variables are determined by a combination of experience with similar products and "tweaking"--making small adjustments until the proper combination of settings is found and then documenting the results.

These might not be the answers you were hoping for, but if that's the case the problem is probably with the questions and your general approach to the problem.

Besides my main question, I also have two subquestions if I could ask...:

1. Is 50 bolts per day per one machine good daily amount? If yes: what should be relation between good and bad parts? I was thinking about 35 good and 15 bad. I think its a good idea. May I please for your comment, what do you think?

Whether or not 50 is a good number depends on the number of parts produced. You will want to have a number that's statistically significant. You say that you've got Wheeler's Understanding Variation, so you should be able to get an idea of appropriate sample sizes from the information therein. As far as proportions are concerned, there's no point in trying to get 35 "good" and 15 "bad"; the object should be to learn about the variation in the process, and how many defectives the process is likely to produce in its present state.

2. I work daily from 7 AM to 3 PM which is classical working time for almost everyone. One of the sorting criterias will be based on worker (first + last name) who is responsible for machine. We don't have many workers so this should be a problem BUT another group of them start working from 2 PM to 10 PM. The only exception is the section for thermal
processes who also work the night job (until 6 AM). Issue is obvious - I cannot be here 24/7. Sorting the cards based on specific worker should also include sorting based on time when they work. Am I absolutelly required to still sort based on work? Note: I have in purpose to choose the bolts for analyzes from different locations of the same box which
says a lot. I ask this because workers' productivity and exactness is not always the same in the different working times (regarding exactly the same person).

[My emphasis]
You need to identify the variables that contribute to nonconforming output and control for them. This means that you should be collecting samples from each shift, operation and operator at the least. You should also know who is setting up the machines if it's someone other than the operator, so you can understand how setup might be contributing. Operators should keep their output separate from that of other operators, at least until the output can be sampled. You might need to use a tag that shows the shift, setup person, operator, and quantity produced.

As I suggested a couple of times already in this thread, before you do any of the above you should understand (by way of identification and classification) the types of defects that are being produced. Although you might be pressed for time at this point, there might be no way to condense the work to be done to fit your available time. There are no simple answers, in other words. The laws of physics can't be altered to fit your academic schedule, unfortunately.

 

[Your Soap]

Jim regarding those specifications limit; while checking the book that you suggested me I realized I was wrong when I mentoined that specification limits are determinated with drawing tolerances (the ones that are visible on the drawings). I see thats actually not true what I was saying. The limits should be determinated through doing a capability research (graph, not sure if are equations required also) and also through cp, cpk, cm, cmk. For each graph function is only one machine. Therefore a lot of functions will be made but I think I can still include several machines on the same graph. Of course I am talking about a graphs for determinating the capabilities and the final result shows what I was asking about the limits.

I already got the software InfinityQS. Does anyone use it here?

I haven't found anything in the book Understanding Variation that could help me for the specific topic about determinating the number of bolts that I need to take out daily. The number 50 was chosen based on nothing, perhaps only the maximum number which I could still handle per one day because of huge amount of machines. So I might be very wrong with this number. Will still need to take care for this.

You might need to use a tag that shows the shift, setup person, operator, and quantity produced.

I will have that datas. They are near each machine. But what worries me the most is if I will have ''access'' to the good and bad bolts, that were produced one day ago, next morning because most of them (good ones) wil already go to the rolling to create the threads unless the customer require to NOT having them. In that situation I should go to the machines for thermal processes.

I apologize for bothering and being so much annoying. I would just like to figure out what am I supposted to do before I begin with anything. Also there is no software that could do a work instead of me, if I don't understand what will I be doing. Will resume with