Standardization: Work Instructions & Target Times

[vonsteimel]

Greetings,
I work at a small mfg company making unique recreational.etc vehicles. We employee about 20 people, 75% of which reside in the factory either doing fabrication or assembly. We've output a total of 60+ units in the year of 2013.

First off, I've posted a snap shot of a couple costing graphs of what is going on with ALL the parts in our factory. The factory is pretty much allowed to run itself and operates more like a fabrication shop rather than a mfg factory. We have, up to this point, used only engineering drawings. This meant, as with most fabrication, it was the workers idea to decide how to make the part. We have some simple tooling such a templates & marking patterns.etc but no real jigs & fixtures like you'd find in other factories. We typically run batches of about 30 - 60 parts per job. This means a guy might only perform the production run for a given part only once a year or longer, making it more difficult to get really good at any one thing.

So this in short hopefully explains the large variation in the costing graphs. Maybe one job was done Joes way and the next was done Johns way and maybe the next guy had a great hunting story to tell which ends up making the costing graphs look like a seismic reading. Now short of finding some new people, we're looking to improve the situation by implementing the next step in our mfg journey. Standardization.

We have the means to implement such "policies" and things may not be as completely out of control as it sounds. Nevertheless, the costing graphs tell a story of their own.

So we're planning to do some time-studies and come up with standard Target times. I think we'll also have to create the work instructions to show how those target times apply. (i.e. cut all the parts at once and then drill them all at once; don't cut & drill, cut & drill, cut & drill.etc) I'm looking for some advice on creating these WI's; we don't have the money to invest in any software so I'll be using MSWord and/or ACAD.... Whatever we do, we're fairly under-staffed in the engineering dept so its got to be relatively quick.

I'm trying to decide how much detail is generally needed. Most of the WI's I've seen are very basic: (i.e. assemble shaft, install clamp, install cover.etc) Is this generally how WI's are? If so, where does the more detailed instructions go as to how to assemble the shaft, how to install the clamp.etc See my attached drawing for some of the illustrations I was planning to use in one of our first WI's. (note; its not in any order but will be pieced out when they're used in the WI); what do you think?

I plan to use pictures instead of sketches to save time, where possible. This seems pretty common?

I'd really like to see some good examples if anyone has any.
Thanks,

VS

 

[vonsteimel]

The illustrations...
VS

 

[KENAT]

1. Do you need full blown step by step work instructions? Would simple written routings with the odd production plan/sketch work well enough for you low volume and the skill level of your staff? I'd only go to detailed work instruction if there is a big time/cost benefit in doing stuff a certain way (which you imply there is) or for stuff that is difficult to fully document in traditional engineering drawings.

2. Keep it simple, since the Work Instruction should only be defining process not the end item they should be easy to modify as people come up with better ideas - call it Kaizen if you must. I'd consider starting minimalist to begin with and only get more detailed if required to get the desired end result.

3. Detailed work instructions may be needed when you don't have very skilled staff or generally accepted workmanship standards etc. For your volumes and your business case you need to determine if it's better to pay a little more to have more skilled assemblers that need less instruction or if it's better to dumb production down as much as possible to pay the workers minimum wage but at the expense of massively increasing documentation overhead/needing to hire a manufacturing/production engineer...

4. Photo's can be fine just make sure they are of a reasonable quality so it's very clear what's going on - annotate them if need be etc. I've seen too many long complex WI with lots of bad photos, some of them showing parts assembled out of order or nut updated to the latest build standard...

5. Get the input of the folks on the floor that currently make the parts more quickly with less scrap etc.

6. Someone may post on here about the benefits of 'going lean', while I still have trouble understanding how some aspects of it really work long term for low volume manufacturing etc. doesn't mean you shouldn't keep an open mind.

(Based on the limited info in your WI sketch, why do you not rotate the template 180° each time to reduce scrap? Add allowance to the template if need be etc.)

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[MikeHalloran]

A while ago I was working on blood cell counters, which contain a lot of small scale custom hydraulics. One of our biggest challenges was erratic seal life, which we traced to a polishing operation on a stainless steel shaft.

Since nobody, including especially me, knew how to do the operation better, or 'correctly', I asked the production guys to:
- Always do the operation the same way. I left them free to choose the way, and asked them to write it down and talk to each other, so that every one of them would duplicate the process, no matter how good or bad it was. I explained that we would keep track of the process' performance by monitoring field returns, which were available in statistically significant quantites.
- Date code the parts, and initial them, so we would know who did it and when, and could thus look up in the workers' notes exactly how.

Pretty quickly the return rate stabilized, and started going down as we adjusted the process.
... but I think worker participation in making decisions, or at least the perception of participation, was the biggest positive factor, as in the Hawthorne Experiment.

Then the unthinkable happened; the company was sold, and moved.
The workers were not moved.
I asked for their notebooks, and they were gracious enough to comply.
I really miss that crew.

With the help of the notebooks, we eventually got a new crew up to speed.

Think of the notebooks as process instructions written by the people performing the process. They were not in any way 'arty', but they were clear enough.




Mike Halloran
Pembroke Pines, FL, USA

 

[vonsteimel]

The material is aluminum Angle so by flipping it, the angles would be on the wrong side of the upright.

Some of the employees use their own notebooks here as well. The problem we have is that their more apt to use their notebook than the drawings.etc so when a rev is made, mistakes can follow because the notebook is essentially uncontrolled. For this reason the owner has largely been against personal notes.

Some gladly give their notes to us, others not so much. The biggest problem is when multiple people have conflicting notes/idea on how to mfg something. Which is a contributing factory to the costing graphs above. Trying to consolidate the ideas can be difficult. Its not uncommon for designs to take on a life of their own as a result, to some extent.

To me, its difficult to keep the WI's basic. I have both a drafting/design degree & an engineering degree. I guess in the drafting school the idea was pounded into to my head to document everything. However, I think KENAT is right. We dont have the manpower to produce highly detailed WI's for everything anyway. I'll start with very basic WI's to get the system going. A watery WI is better than no WI I suppose. We can add more detail as we go along if need be...

Where/How do companies generally record the equipment requirements for a given operation? Such as hand tools (pliers, utility knife, tubing cutters, 11/16 open end wrench.etc); Power-tools (cordless drill, die grinder.etc) machinery (bandsaw, lathe.etc) and safety equipment (safety glasses, ear-plugs, cut proof gloves.etc)? Or do they?

I'm trying to integrate it into the WI's as a way of reducing the "gap of interpretation". i.e. "Cut material as specified using bandsaw"...
Thanks,

VS

 

[racookpe1978]

Do you repeat the same build 9geometry and arrangement each time, or is each of the assemblies "custom made" to "new dwg's and to new "instructions" ?

You are, at least, in the same place using the same workers.

Historically, from ALL repeated projects since the pyramids to cars to Victory ships to tankers to Liberty ships to airplanes, doing the SAME job by the SAME workers in the SAME location using the SAME tools reduces costs: Costs increased even in WWII when the same shipyard changed.
Costs increased in Bethleham-Fairfield at changes from Liberty ships to LST, from Libertys to Victory's.
At California (Kaisers) from Liberty cargo to Liberty tankers, from Liberty tankers to Victory's.
At Oregon from Liberty's to Victory's, etc, etc, etc.

Across ALL shipyards in WWII, the nbr of days from keel laying to launch (one fundamental, never-changing, easy-to-count, very reliable, very basic "cost" to build a ship in both number of manhours AND length of time), went from
175 days (ship 1)
87 days (ship 2)
54 days (ship 3)
50 days (ship 3)
48 days (ship 5)
46-40 days (ships 6 to 10)
40-35 days (ships 15 through 20)

BUT - When the same company built a new yard with new management and new crews? When a "new" type of ship was changed but worked under the old crews in the old yard? When the same ship was started with new crews or even under the same management in a new "addition" to an existing shipyard? Back to ship 1 or ship 2 times!

If you are content with your current costs and current slow construction times, keep your "old" management tools and training and tools and jigs and crew attitude. Keep the design the same (or fix it!) and improve your craft "in-hands" tools and jigs and "smarts" - (maybe by changing craft supervision!) - and you should see 30 to 50% lower costs.