Work Instructions - How detailed do work instructions need to be?



We are a machine shop currently seeking QS9000 certification. We had our first pre-assesment audit on August the 8th. We did not have any majors, just a couple of minors and observations. You know, a little word play :) Anyway, our registrar did not even view the first work instruction! I'm sort of worried about this. Let me give you some background. We are a small machine shop that has both CNC and Conventional machines. For work instructions, I gave a "general" way of using each machine, but placed the majority of the work instructions upon the shoulders of the original machine manuals. The machine manuals are now controlled documents and a part of our work instructions. I think that this will cover the "operations" work instruction area. What I'm asking really is, how detailed to work instructions need to be? Would could happen if I'm missing a couple of work instructions during the compliance audit? Do I need to go as far as to write work instructions for the use of forklifts and pallet jacks? I really need to know to what extent a work instruction needs to be.

Please assist with any information!

Joe W. Guy,
QS9000 Administrator

Kevin Mader

One of THE Original Covers!
How many work instructions are necessary or required? How many times I have heard this line? Plenty.

Lassitude is correct, more of a training issue me thinks too(hahaha). A well written job description will cover most of the basics of operator requirements (i.e. must be able to run and maintain an XYZ machine safely).

Here is a test you might want to try to determine if a work instruction is advisable.

1. What is the risk involved with the process? You might want to use a Risk Analysis tool to answer this question (read a post I just left in the FMEA and Control Plan forum here at the Cove).

2. What would happen if a work instruction did not exist? If you think that the absense of a work instruction would have an adverse effect, create one.

3. Have you experienced repeat failures in an area where training has been done with little positive effect? Perhaps create one here or investigate the trainer or training methods (perhaps do that first!).

My advise, don't over burden your documentation control person with useless (nonvalue added) instructions. Keep your program neat and clean, and most of all, relevent.

I hope this helps. Back to the group...


Molissa Burton

I agree with all of the posts about training. If you show evidence on the training records that a particalur employee has been trained (or has a certifcate) to drive the forklift, then you don't need to post a work instruction on the steering wheel. As far as the machines like the CNC you are going to want to train them on how to use it anyway, at least for safety reasons. So you can leave the detail to the operators manual. You will probably want some vague instrutions, whose responsible, who can run the machine, what to do with nonconformance, etc.
Realisticly, if you had a questions on what really goes on with the CNC you would probably turn to the manual.
Just remember, the more training the less the paper (most of time anyway.)

Kevin Mader

One of THE Original Covers!

Sorry to get back to you so late (I overlooked this forum). Thanks for the kind words, but I must admit that my post takes from many of your own posts. So thanks for the insights.

The post in particular that spelled it out the best for me was the one where you commented on doing FMEAs for processes (operations) as opposed to running through a Process FMEA for each product (QS9000). Made excellent sense to me, as I have found in my experience that the past FMEAs we ran were quite redundant. A great time saver once you get over the initail investment of doing FMEAs on the determined CRITICAL operations (don't get hung up on the insignificant). But I made an additional discovery.

I have been working with our Chicago facility to help bring them to ISO 9001 registration. A common question asked was to the point of when to create a work instructions. A review of our past FMEAs indicated that we planned to control everything through the creation of work instructions. There associated RPNs were for the most part, rather low and part of the "trivial many". Why create work instructions where it has been decided that a specific step in a process was of a low risk? In lieu of this discovery, instructions are only created where risk is determined to be significant (reasonable risk). Now, instructions are issued less than 1/10 the time they once were.

Aside from applying the FMEA process to the manufacturing environment, I have instructed other in the use of this process and encourage them to use it in their worlds (all aspects and departments of the organization). The advantage to doing this type of FMEA is that it creates a record of our thought process. Since folks are the authorities in there respective areas, they give due consideration to the activities they have identified as being critical to the daily operation of the organization. They will invite their internal suppliers and customers to these meetings to fairly represent the overall needs of those involved. What I have noticed, folks often over complicate issues, often over/under estimate impact, and create extra work while individually estmitating process inputs and outputs. The group effort, however, tends to ground things better, they debate ratings in a collaborative effort. The side benefit to this process, increased,better communication!

Determining risk involved is essential in my opinion. Use a risk analysis tool and use group concensus. Often times, folks just don't know where to start. I was one of those folks too! No harm, no foul. Hopefully, people visiting your forums can take away from the experiences of the many folks here who, in one way or another, fell on our noses as often as we have hurdled seemingly insurmountable obstacles. My suggestion here is just one of many solutions to this problem. I hope it helps.


Top Bottom