determining time and cost in a machine part

look up 'Time Study' for your answer. Sounds like your company is guessing.

A good time study uses a stop watch and at least a 30 pc run and averages the cycle time.

as for set up? same thing, but before you start that watch, make sure all the unusual items are out of the way first, like all the tools are where they belong, all the fixtures are in the right location...

Why would writing done the performance-times for set-up and machining tasks give errors? What are the sources of the errors in the way it is currently being done? May be you just need to eliminate the sources of errors.

Estimate the time with consideration of the complexity of the part.

Most machine shops that I deal with have set up fee that is tacked on the cost of the part. Its a set fee. So its the cost of one part plus the set up fee or the cost of ten parts plus the set up fee.

I don't think that they go and look at all that needs to be done to generate a price. They generalize the time and charge a fee for it. What they may loose on the set up on one the make up on others.

I the present economic environment your question leads me to believe someone is questioning whether it would not be better to out source the parts. My question to you if the parts that you make in house will take a week to get. What will that extended time cost the company. In a manufacturing environment the losses from a single downed machine awaiting parts would well pay for the man power for the year and machinery necessary to make the part. The cost of inventory of all the parts would also be absorbent.

Contact Precision Machined Products Association in US.

www.pmpa.org

Matt Dougher 440 526 0300

They have a "Development of the Process Hour Cost Rate Manual" that will help you set up a logical chart of accounts.

Most shops do a quote or estimate, then compare actual to quoted. Then work the hell out of the gap. Once you have a standard then you can do a gap analysis.

High volume contract shops use software to manage this down to the third decimal point. PMPA can give you leads of software companies being used by their members too.

In aggregate, you can then see how your costs accrete, and where variances are.

They also offer benchmarking services in the financial area.

Even though you are a captive shop, you could become an associate member.

milo

let me address this a bit further.

What we have problems with are the different processes involved. For example, They may shear enough raw material to size for 1000 parts. But it will then sit in a bin for weeks has it gets used from the brake, to the CNC mill etc...

The errors occur because the "machinist" we have do not accurately document the time and because each set up is different and requires different tooling they sometimes count the time it takes to gather the jigs, tools, and material and sometimes not. They try and use military time but cant get that right either.

What I was looking for was some sort of automated or "system" that others use to reduce this to get a more accurate accounting.

What we don't want to do is outsource! We have 4 CNC lathes, brakes, shears, etc. A full machine shop! And we definitely don't want to have to have a manager sit with a stop watch for every single part/process.

The other thing that I am thinking is that the time it takes to do a part is really irrelevant is it not? Because there are so many factors that are uncontrollable. Wouldn't it be better to just for one day have them work on one part start to finish that includes all the different processes and then divide the number of parts by total hours to get how many per hour? Then multiply that by the shop rate?

Guessing is the big problem I am not sure anyone really knows what the shop rate should be. In fact we have just recently found out we don't have raw material prices in our database and nobody seems to know how much to mark anything up? its insane!

Surprises me to this day they have been in business for twenty years!

bill12780

Your remark:

Wouldn't it be better to just for one day have them work on one part start to finish that includes all the different processes and then divide the number of parts by total hours to get how many per hour? Then multiply that by the shop rate?

Tells me that you are thinking in line with Dedicated processes and production Leveling. The Japanese are a master at this and even wrote extensively on these concepts. So you may look into some such books, including the concepts of World Class Production processes. Complement the production methods with the production leveling techniques and you should be alright.

Tell them how many different part numbers you might have to make in amonth Bill.

Bill --

Seems like you've got a nice mess in your shop. Listen to Milo. He knows what he's talking about.

How soon we forget. Back in the days of union shops and incentive pay systems we had cost accounting and time and motion study engineers and standards for set up and run time for every engineering released part in the shop. It seems like that is all gone now.

This is not a good venue for teaching you all you need to know. My reference books on the subject are all old and way out of print.

But one of these books is just as valid today as it was when published in the 1980's. The book is:

AM Cost Estimator by Philip F. Ostwald. Penton Education Division 1100 Superior Avenue, Cleveland, Ohio 4th Edition ISBN0-932905-06-4. Earlier editions by Ostwald I think were published by McGraw-Hill and may have been titled "American Machinist Cost Estimator"

Buy, beg, borrow this book any way you can. It's out of print as far as I know. (After a year of searching I bagged one on ebay) It is full of information and methods you can use. Some of the constants in it are 25 year old vintage like material and labor costs. The constants for CNC machine performance like setup and rapid traverse times are from that generation. Also cutting speeds and feeds reflect the cutter technology of 25 years ago. These old constants will give you conservative estimates of setup and run time. You will have to adjust labor rates and material costs to today's marketplace. This is a database you should be building immediately.

The methods in the book will give you cost and time estimates. You will still need an orderly system of shop work orders for individual parts and a way to capture expended hours, dollars and other info on each part for comparison to estimates. The gold is in the comparisons because they lead you to the discovery of inefficiencies. How you deal with the inefficiencies will determine whether this tool saves you or causes your demise.

The cost estimating calculations can be reduced to an Excel Workbook format once the manual approach to calculations becomes burdensome. It's a bit of a project. It should start with your biggest machine tools or work centers of machine tools. You'll find that the CNC machines are among the simplest calculations since they have only a few basic moves and cutting rates for each class of parts, pretty much standardized tooling operation methods.

I know what I'm talking about here. Two years of my engineering career were spent working with this tool, a complex Excel workbook and refining it in the real world of semiconductor equipment parts manufacture.

A warning here. Systematic cost estimating for manufactured goods is an incredibly powerful tool in the right hands. Politicians from within and outside your organization will fight it tooth and nail and mount all kinds of challenges. You ready for that?

Ed Weldon

Bill

My advice would be to slow down. Way down.

Start with the assumption that your coworkers are giving you 80 to 90% best efforts based on the systems in place. They probably are.

Figure out which systems will cost the least to change. Then figure out the systems that will produce the most benefits. Don't use just your opinions ask the people who know. Do the cross product and determine which "fixes" have the lowest cost and highest benefit. Then, ask the people effected, "Do you think it would be OK if we...." Most of the time they will think its fine. Sometimes, they will know things you don't. Sometimes, you only find out that they don't like change no matter what. That too is important information. An indication for change, in my opinion.

Think about how the people you are working with feel. Respect them.

Would you like your boss to spend his time measuring your performance with a stop watch for a half hour? Would you do your best if you knew you were being rated?

Most of the data you want is in the machine controllers memory. You just need to learn to get it out. No stop watches, no time studies, just accurate data. Won't do much to improve things though. Just gives you an accurate measure of the madness. It will give you a baseline to measure change.

To make meaningful change you need to lessen the amount of wasted time between spindle on for last part of "A" and spindle on for first part of "B". To do that you need to create a flow. With a small shop like you talk of that means when the brake operator finishes bending he delivers to the mill. The mill delivers to his customer and so one till the finished goods are shipped.

The key to making that happen is a good Router. A production plan sheet that goes with every part, plus shelf space right at each machine with some over capacity for the area.

Then work on getting all the tooling and fixtures and inspection equipment organized. If possible label everything with a code that tells everyone where "home" for this item is and get them trained to return send it home when finished with it.

Then look at reducing work in process and inventory by improving your companies shops ability to turn over an order. This ability to get things done in a timely manner is far more important than the seconds per part per operation.

Foremost, be thankful you are inheriting a small problem. When McNamara took over production management at Ford he found they hadn't been doing basic accounting! In order to estimate the dollar value of accounts payable he had them weight a stack of invoices, and then weight ALL the invoices, Then they ran a tally on the sample stack which came out to just over $100,000. From this they were able to estimate the accounts payable to be approximately 45million. Turned out to be a pretty good estimate when all the accounting was caught up with.

You just have to figure out how to network a bunch of machine controllers and interrogate them about what is going on while earning the respect and goodwill of you fellow employees. Listening to people and buy doughnuts are great ways of winning friends and quieting enemies in a machine shop. Running around with a clip board and a stop watch isn't. Insanity is par for the course, don't mess with the American way!

Best wishes,

Mr. Gee

You need to get a few things accomplished...

First of all, you need to do a time study of each phase of your production..how many parts produced an hour?...how long for set up of new job? (Which by the way will fluctuate...so you need to not only compare set to set...but also job to job (bc some transitions take longer to accomplish due to more tasks involved. The more data you are able to gather, the better you will be able to note differences. Every task associated needs to be documented including task performed and amount of time.

You need to evaluate your scrap associated with production, as well as how much your overall labor costs are. Can the scrap you produce be reused, or is it a marial loss? You also need to compare your quoted hourly rate to your rate of yield..yield being the production numbers minus losses.

With all of the data collected, sit down collectively with all people associated with the job to analyze if there are ways to cut your operation down for efficiency. Are there unneeded actions? Can some steps be combined? Is there a better approach/ technique? There are generally alwas ways to improve on a production process...whether through production improvements or through maintenance/ mechanical corrections.

Hope this helps...

Garrett

Hi All:

I have enjoyed you answers but I was surprised to see that no one suggested using a camera to record the set-ups to save time the next time the job is set up. This saves the operator figuring out what is needed and how it goes even if they have a printout of what is required. Also, years of experience makes a difference in how many time an operator has set-up a job or simular jobs.

Just as a side note I have been estimating how long it will take me to do a job at machining for 57 years. Just last week I got caught by the extra time it took to unwind the rats nest of UHMWP3 chips off the short parts I was turning. This took longer then the machining. Sooooo you never really know.

GOD'S BLESSINGS

Bill

If you freeze the plastic before machining, it should help embrittle the material, resulting in shorter chips. Give it a try.

milo

Bill -- I already flew in here with a lengthy description of a specific tactic (cost estimating). On reflection I should have started with strategy and the big picture.

The real question to ask may be what kind of management does the shop have? If you have the position in your company to make those observations, the managerial knowledge to analyze them and the respect to have them listened to and acted on then you are good to go. If not you are on a fool's mission.

You are being asked to do what the shop foreman or superintendent should be doing himself. At the very least you should be acting under his close guidance and approval. If you are not, then there is something very wrong and YOUR management is allowing it to happen.

There is a very important information feedback system in that shop that isn't working. By your own description it is producing inaccurate data. Effective management is a closed loop control system. Plan, execute, observe results, learn and fix as needed. Repeat. etc.,etc,......(I'm assuming here that the shop is big enough that it actually needs formal management. With 4 CNC machining centers it sounds like it is.) The shop boss may be allowing all his people to plan and execute their own work. In a small shop where he can observe the results directly and carry it all around in his head that can work. As the boss's tasks grow with a growing enterprise he may find he no longer can do that. To the extent that outside tasks command his attention the primary mission of production goes open loop. So he creates a "management information system" rather than hire or appoint additional foremen or leads. An MIS is a poor substitute for personal attention; but under the right king of leadership and with thoughtful crafting of the method it often works.

So your first task is to put the current system under a magnifying glass and find out what isn't working and why. While it is interesting to look at alternative methods a new method may well take far more work to implement than fixing the old method. And don't forget that the source of the problem may have little to do with the method of data collection and everything to do with human attitudes and emotions and the real leadership structure in the working group. This is where any level of respect and credibility you bring to the party will count for a lot.

Ed Weldon

Thank you all for the very insightful replies.

I am sure that the information here will help me to steer the managers that are involved.

I would like to make sure that everyone understands that we are not blaming anyone for the woe's of the machine shop. Everyone knows that its not the people down there. Its the system and know one person is getting blamed.

Our problem IMHO, is growing pains. It was only three or four years ago this whole place was ran out of a storage unit by five people. Today we have 50 or 60 employee's and two location nation wide.

Has the company grew the systems did not. And that is what is causing our problem. Our shop does good work. But nobody knows how much it cost us to do so. And that the real problem.

Thanks for all the input I will condense it and give it the guy who asked me to do the research. Hopefully they will make some good changes and things will get better.

best to all

bill12780