Lean Manufacturing

[Guest]

.. I just Hate sitting around idle sometimes

I LOVE "idle" time. This is the time I use to get in some training and complete "backburner" projects.

[Guest]

I often see the problem of too few tools in the shop. I get that the bean counter sees an $800 tool assembly on the shelf and gasps. But if you're having to rob tools from a machine that is down to run another job, you don't have enough tools. It's an expensive business.

[Bob W.]

If both machines are busy and the new machine with 133 tools was running job A and B and the old machine was running job C, when job C was complete either job A or B would get moved to that machine, tools and all so they could be run in parallel. No way would I load all new tools into holders if the they were already set and just needed to change machines. I figures 133 tools would add some convenience, but not enough to justify the cost and I wouldn't be able to tools permanently for all of my production jobs. With 300 pots or so I could do a fantastic job of that but I don't have enough space or the budget to swing that.

 

Tool management is by far our weak point right now. Sourcing and loading tools consumes the most time by a factor of 3X when setting up projects. The programs are always on the machine, the fixtures bolt to the tombstones in minutes, then the tools take 30 minutes to source, load into holders, and load into the machine. Most of the time we are able to get most of this done while a project is running but not always. Fixing this is high on my list. I currently have about 300 holders and I add about 6-8 per month.

 

Any ideas on how to streamline the tool setup process?

[Guest]

Well, sometimes the "best" tool for a job may not be the best tool. For example, I will often interpolate a hole rather than drill it even though a drill will make that hole faster. Single point tools instead of taps. If I can threadmill 2, 3, or 4 thread sizes with a single tool, I've just saved myself a holder, setup time, etc... Obviously a high production job needs to be optimally tooled to make the anticipated profit, but short run jobs afford you the opportunity to get creative.

 

JM2CFWIW

[Newbeeee™]

Bob,

What we did was to get a couple of roller door cabinets, and made some custom shelves.

We got a fab shop to make some shelves that would fit inside out of 2mm sheet steel,, and punch some 35mm (can't remember) holes to take the tool holders in a grid pattern.

We can get 40 holders per shelf ish (I could take a piccy if you want).

We then got them powser coated and pop riveted the things permanently into the cabinet, so the whole thing is rock solid.

What you could do then is to leave tools set on the shelves.

And if it only takes you 30mins to get tools together and into holders etc, you're doing very well.

[JeremyV]

My one co-worker who'se a programmer had a discussion on this very topic with our supervisor.  He of course went on to say how we don't do lean manufacturing to where the supervisor said we did.  Of course it went back n fourth and so on. 

 

It wasn't till the co-worker compared doing yard work and by using that as an example:  "Say you have 4 sides: front, back and sides of the yard;  do you mow the front yard, stop grab another tool to do trim work then cleanup and some other thing then move onto the next side and so fourth or do you mow the entire lawn then do trim and followed by cleanup?"

 

Kinda amusing really.

 

And @ Bob, we have that issue here too.  Across the total CNC machines in the shop and this is just the # of pockets in the machine: there are 270 total pockets, so we have at least 350 tool holders give or take 50 or so, but those 50 aren't being used very often and aren't collet holders.  Theres enough for maybe 10 laying around to steal off the shelves where other machines sit at.  Believe me, I would have loved nothing more to have plenty of tooling sitting around to be able to do tool setup for the next job or keep standard tooling sitting in the collets and that way setup would be a breeze.

 

Hell, if it were my decision, i'd replace 4 machines in the shop: a boring mill that is hardly ever used anymore and its just sitting there looking like a lawn ornament, an aging SQT and dual turn + mazak vertical.   What i'd have loved to do is get a newer dual turn, a multi-axis cnc lathe, and a 5 axis mill, inaddition to that, another horizontal mazak.

 

I'd commit the old horizontal to running production on our line of products that require a lot parts   IE the 8 piece sets and the other horizontal to everything else.  I'd have the flexibility to not tie up one horizontal and bring in new products that require a little more complex milling / turning.

 

Our company in general is more of a custom shop that specializes in vacuum chambers.  Less than 50% of chambers are the same, a lot of the times the design changes.

 

 

And as a programmer, I try to streamline tools as i said before.  But it does make it challenging to decide on whether to use the same tool or switch over and add a larger one in its place. Sometimes there is nothing I can do to make it any better than it is.

[chris m]

 

 

It wasn't till the co-worker compared doing yard work and by using that as an example:  "Say you have 4 sides: front, back and sides of the yard;  do you mow the front yard, stop grab another tool to do trim work then cleanup and some other thing then move onto the next side and so fourth or do you mow the entire lawn then do trim and followed by cleanup?"

 

I guess I am dense, but I don't understand the point of this analogy. I don't think anyone is advocating bailing on a job that is 1/4 done.

 

C

[Bob W.]

From my experience I try to fill my customer's current order and that is it. I might run an extra cycle or two just in case the anodizer loses some parts or one of my employees drops a part (or two...) but I do not want anything sitting on the shelves. It takes up space and it is a liability. I have had times when there was plenty of inventory sitting on my shelves and the customer made a design change. All of the inventory went in the recycle bin. The goal is to be nimble, to set up quickly, to tear down quickly, to move the project to another machine quickly... This makes it easy to jump when the customer says jump.

 

My analogy would be this. If the customer wants a tour of your landscaping and they will have a look at the front yard, then the side yard, then the back yard, then the other side yard, the mowing would be started a short time before the customer arrives and each area they looked at would get done (to completion) right before they were ready to see it. It is all about the customer.

 

The most efficient way is the cheapest way over the long term but probably the most expensive in up front costs.

[Joe788]

My analogy would be this. If the customer wants a tour of your landscaping and they will have a look at the front yard, then the side yard, then the back yard, then the other side yard, the mowing would be started a short time before the customer arrives and each area they looked at would get done (to completion) right before they were ready to see it. It is all about the customer.

 

 

We don't let them see the side yards - because that's where we hid all the garbage from the front and back yard.

[Guest]

NEVER let the customer see the kitchen. NOBODY needs to see how the sausage is made.

 

[Matthew Hajicek - Singularity]

IMNSHO "Lean" is often taken too far.  Sure it saves money to not keep parts on the shelf, but it also costs.  It costs by adding lead time and complicating scheduling.  What happens when you get all of your customers asking for stuff all at once?  Either the lead times add up (which will make some customers turn away), or you have to job stuff out, or you have to turn jobs down.  In my experience this often means everything has to be made all at once at the same time, and it's the fault of the guys on the floor if they can't keep up.  So in reality you should weigh the chance of each job with the cost to run it now vs later vs what's vying for machine time.  It's never a simple answer.

[chris m]

That is correct, sir, the answer is rarely simple; you have to build your own framework of what is reasonable, and what isn't.

 

C

[metalmansteve]

I think the idea of lean is great but sometimes common sense needs to prevail.  Back to the original post....you have an order for two parts but one part doesn't ship until next month.  Running one, tearing down and setting up again next month doesn't make any sense.  Maybe if the setup time is 10 minutes and the cycle time is 10 hours but for most jobs it seems obvious that you just run one more part while it's setup.  Putting parts on the shelf and shipping from stock has worked extremely well for me but I do understand the argument against. 

[oneyankfan1]

True cellular manufacturing in a machine shop has to do with single part flow. To obtain this you must be able to implement S.M.E.D. (single minute exchange of dies) also you must take all "traditional" setup factors out of the equation. No edge finding or sweeping in holes, and all tool magazines must be standardized. Tool setting is normally done on a presetter. There is a shop here where I live that has it down to an art, Cox Machine, their entire facility is truly lean, they look at all of the lean tools as building blocks and if you remove one the whole house will fall down. Not to mention it is probably the cleanest shop I have ever seen, thanks to their HUGE emphasis on 5S.

They are impressive, all operators are certified to buy their own parts, using faro and roamer arms, which drastically streamlines the inspection process. They use the same process on all parts even FAI's. Also they have recently looked into additive machining for all fixtures. I worked there before I was a programmer, and they have grown a lot since.They also apply this to their assembly department, and office.

 

http://www.coxmachine.com/

 

http://www.youtube.com/watch?v=ZorHL4iM-WU

 

Just my opinion but these guys have it nailed.

[Patternmaker706]

Batching makes more sense than just in time.... The guys that come up with this crap are clueless. It worked for toyota 40 years ago and every one thinks they can duplicate it.

[Guest MTB Technical Services]

Batching makes more sense than just in time.... The guys that come up with this crap are clueless. It worked for toyota 40 years ago and every one thinks they can duplicate it.

 

 

One of the most ignorant statements ever made here.

Someone who doesn't knowing anything about the subject pontificating about what actually works.

Deming was a Genius.

 

Just in Time has been PROVEN to work.

It is NOT a theory.

 

You clearly don't know what JIT, Kaizen and Lean Mfg actually are and obviously have never actually been exposed to it.

 

One interesting point I've observed regarding JIT, Kaizen and Lean, the people with the least experience in JIT, Kaizen and Lean

like making parts that don't have any customers and just sit on shelves.

[Patternmaker706]

I've been to you seminars. We've sorted, set in order, shined and those other S's. Ive spent thousands on kaizen events. Nice dog and pony show! But when it takes hours to change a machine set up for one part because the "software said so"... Makes no sense asset all!!

[Patternmaker706]

Just my ignorant opinion...

[Guest MTB Technical Services]

I've been to you seminars. We've sorted, set in order, shined and those other S's. Ive spent thousands on kaizen events. Nice dog and pony show! But when it takes hours to change a machine set up for one part because the "software said so"... Makes no sense asset all!!

 

 

Like I said, You obviously don't know anything about it.

Here's a hint, If your setups are taking you hours, you are NOT using JIT, Kaizen or Lean Mfg techniques.

 

Stick to particle board.

[RaiderX]

Lean has a lot of different aspects (techniques) not all fit in any given application but one would be wise to apply what fits.

we run mostly 1-3 qty very complex part onetime orders and cannot afford to tool plug and play for Jit but do use many lean principles.

standard tools, standard fixtures, storyboards etc.

it does bring value

 

Doug

[Newbeeee™]

Okay – story time with my full heavy duty industrial flame suit on

The below is my only real-world experience of lean manufacturing working for someone.

 

My last job, I was Engineering manager of a tier 1 supplier to aerospace.

We always had cash flow issues as the owners’ son who ran it, really couldn’t run a bath (he was out of his depth in a puddle J). He had purchased a load of Hitachi machines ($2mill boat anchors) and moved factory from their owned one to a rented treble the floorspace when the company only needed a small portion investing. But he wanted shiny shiny new and didn’t care about the monthly repayments…

Anyway, because we were always in the can, our deliveries were sh!te because we never could pay treatments houses on time so parts were always back late.

Our major customer (80%) was a well-known aerospace company that had the same name of a tyre making company…

They were peed off with our deliveries so they offered help and sent in their own internal consultants.

We had only multi-axis work running through 3x mill-turns. Most was their work, but some was other customers.

Materials were varied – stainless, brass and ally, so there was quite a bit of swapping over with tools/tips etc for the materials. That said, not much down-time in the scheme of things.

The 1^st thing the consultants wanted to do, was to convince everyone in the place that we needed single piece flow. We wouldn’t then hold stock (not that we overmade as we could only afford the material for the job we were making), would only make what they needed when they needed it, and as a company we’d be quids in.

They offered to take us all up to see their internal manufacturing cells as they had been awarded ‘centres of excellence’ from some place that I can’t remember now.

So a few of us went there where their cells had 2x cnc lathes (2axis), 1x VMC with 4^th axis, 1x studer grinder, 1x optical profile measuring machine, and a deburr mop in it.

One operator per cell with a process of:–

Load material in chuck of lathe 1 and stand waiting.

Take part out and load into lathe 2 and wait.

Then to VMC and wait.

Then to grinder and wait.

Deburr manually (10 secs)

Inspect

Then repeat the process (it was around a 15 minute cycle time, which for the operator was a 13 minute waiting time.

So we (management) said it would never work for us as we were a jobbing shop, and the machine efficiency would plummet, and so would output which would bankrupt the place.

So the consultants changed the approach, where we were to only run ‘suitable parts’ on the mill-turns. This meant only stainless parts and taking off other customer parts, and the remaining machine capacity was topped up with a lot of 2-axis parts.

Material diameters were also increased (some majorly) to save swapping out spindle liners (again this didn’t really take much time either) but?

The other customers’ mill-turn parts were then 2xaxis turned, and VMC-milled and lead-time rocketed and we were late, losing one customer.

Less than 6 months later, the place ran out of money and went bankrupt.

 

Looking back now, I still feel that the customer cells with the one operator was a huge waste of money, as the machine efficiency was p!ss poor. I reckon the investment in machinery alone was $0.6 mill per cell, where most of it was idle for most of the time. And they had 5 or 6 cells.

Don’t get me wrong, I fully understand that set-up reduction is king, and that single piece flow can work, and this customer obviously didn’t care about machine repayments – they had an allocated budget for implementation and that was that.

I think they could have been far more efficient by small batching.

 

Tim wadda u thunk?

(excuse the funky text - written in word and pasted in)

[BenK]

 

One operator per cell with a process of:–

Load material in chuck of lathe 1 and stand waiting.

Take part out and load into lathe 2 and wait.

Then to VMC and wait.

Then to grinder and wait.

Deburr manually (10 secs)

Inspect

Then repeat the process (it was around a 15 minute cycle time, which for the operator was a 13 minute waiting time.

 

That cell setup would work great if there was a part in every machine at the same time. Not just one operation at a time. 

[Guest MTB Technical Services]

Tim wadda u thunk?

(excuse the funky text - written in word and pasted in)

 

 

The waiting should tell you something.

Nothing about the process described is Lean.

Waiting is Waste.

[Guest]

I've been to you seminars. We've sorted, set in order, shined and those other S's. Ive spent thousands on kaizen events. Nice dog and pony show! But when it takes hours to change a machine set up for one part because the "software said so"... Makes no sense asset all!!

If your setups are taking hours, you're not doing it right. Look, we've got customers that have 3 turret multi-tasking lathes. 30-40 minute setup EVERY SINGLE TIME! No exceptions.

 

Your company probably needs to take a seriously long and hard look at how they're doing things. If you're a milling house, you should be looking at common workholding. We've got customers that do short run (1 or 2 pcs.) on the HMC's and 5-Axis, "setups" ( if you want to call them that) consist of mounting a piece of material, load a new program (unless it's going on the Dataserver, then the Programmer loads it), attach the program to the pallet, add any new tools to the magazine, set new tools TLO to "99." and run it. When a tool that has 99. in the offset it get measured and the program continues. Easy Peazy! They have even gone so far as to go with common material sizes. Now they get better pricing on material because they are able to purchase larger portions of the mill run. They only add tools an necessary. They keep the tool magazine loaded with common tooling. SOme tools are not exactly THE most efficient to do a job with like if a stub tool would be best, they have a standard because they can use it on multiple jobs.

 

No lie, it's expensive to get started, but you can do things incrementally so the upfront isn't so steep and if you commit to it, you can have things going in the right direction in a mater of quarters... IF YOU'RE COMMITTED. If you're not, it WILL fail, and you'll say "see I told you so", when in reality you still didn't do it right.

 

JM2CFWIW, YMMV

 

[Newbeeee™]

The waiting should tell you something.

Nothing about the process described is Lean.

Waiting is Waste.

This was our thinking. We looked at what they were doing and thought WTF? It didn't make sense.

But we had to be real careful what to say because they had 'won awards' for it. We kept hearing that line.

Kings clothes and all that.