Understanding Scale In Manufacturing Let’s Build

This post is going to discuss the important issue of scaling the amount of product you produce.  Hackaday had a recent post about OtherMill. But before that here’s a short video from Tested.


Tested also has good review of what an OtherMill can do.


This post though is going to look at dealing with manufacturing planning and how not to get yourself in too deep by biting off a bigger bite than you can afford.  This post of a tour of the OtherMill shop floor is a good demonstration of how to do it right, more or less.  while OtherMill could do things a bit differently , by and large they have a good handle on their processes.

Lessons in Small Scale Manufacturing From The Othermill Shop Floor

Here’s another article from Make.


Hackaday also has another post on the Pocket NC which has also bucked the trend and done well by keeping costs under control and not building too far ahead of themselves. No matter how successful you think you are and how many orders you have, if you can’t ship and make a profit, then the exercise is pointless.


How Did Pocket NC Survive and Thrive?

Some people have found this out the hard way.  Leaving disappointed customers and lost hopes behind.  In any case failure before you even start is not a good thing.



So how do you avoid having to rid on ever higher piles of cash to keep things going.  First of all you make the decision on what to vend out and what to do you self.  For somebody starting out I would expect to vend out almost everything.  The reason is that way you can borrow the experience of your vendors. A good vendor is also a good partner. I’ve worked on both sides of that issue and vendors want new business. They also generally want to make things easy both for them and for you.

When vending out a part there are four costs involved. There’s the engineering charge, the tooling charge, setup charge and manufacturing charge.  When you send out a RFQ potential vendors may cost these things out in different ways.  Try to make sure that you get the nonrecurring charges and keep them separate.  You should not have to pay them over and over.  It’s also best to send out as many RFQ’s as you can and plan to use two vendors when possible. Also consider different manufacturing methods and compare.  For instance a machined or a sheet metal part for the same application.  Also ask for escalating quantities for the parts.  I typically ask for 50 for prototype, less than that ca waste time if you are working at a high pace, 100, to see if there is a cost/part advantage even in the prototype stage, 500, 1000 and 4000.  I might go to 10,000 is the part is a common one sued in multiple places.

Make sure you understand the lead time issues that may come up. Lead times can be critical and can really screw you up if you are not careful.  Insert the lead times into the project schedule and work back from expected deliver and add a couple of weeks for padding.  There’s nothing like finding out that there’s an unexpected delay in part deliveries to screw you up.

All parts should be permanently marked with the part and revision number.  If the part is too small to be marked make sure that the parts are shipped with the PN# and Revision marked on the packaging and when you receive the part make sure that they are properly separated into the correct revision.  If the part is going to require secondary operations like plating or finishing, make sure that you take the same steps for the secondary op as well as the primary.

Always include a current revised drawing with the RFQ and with every additional PO.  Each PO should have the revision on the PO.  Have a process in place to make sure that the vendors and whoever’s doing purchasing knows when a revision is pending. Being able to stop production while change is underway can save money in scrap and wasted time.  The small delay doesn’t usually matter unless the change is significant.  In that case releasing the change should be done at the beginning of a new order cycle.  On your change form, make sure that you include instructions on what to do with existing inventory.  You don’t want parts that should have been reworked or scrapped get into product.

If it make sense space out the production to more closely match your orders.  Sometimes that is possible and sometimes it’s not. In any case it works better to be flexible with vendors on scheduling especially if  you are startup.  They can fit you in between their larger and more important customers.  Frequently this can work to your advantage when delivery comes in early and they might cut you some slack on price so that they can keep their shop running.  At the early stages they will probably insist on this anyway, but plan to POD and not on terms.  I know that having the money around is important, but you are building good will which can help you when you need something expedited or when cash flow may BE an issue.

Plan for assembly and look to make things as easy as possible.  When yo are prototyping try different assembly method to see how well thing go together. Keep an eye out for those gotcha like a screw that needs to be adjusted while peering down a deep dark hole.  On a fragile part that breaks.  Take every assembly apart and put it together again in  a different order. Design parts so that they can only be assembled correctly.  A tab or pin, an offset screw hole to guarantee orientation can save time and reduced errors.  Make all the wiring in one harness that drops in. Make sure that the right connectors are only able to go into the right pins on the boards. Use connectors with additional pins in them as dummy pins to make sure that only the right box connector fits into the right connector on the board.  Use different type of connectors if necessary.

Have spreadsheets for planning for different production levels with the numbers from the quotes you have in them. make sure that you get production times for the key components This should give you a fairly good summary of what to expect for any production size.  Pad those number 20% or so when doing the estimates for the problems that will come up.  Work up a Gant chart for the production schedule starting from the expected ship date. Make sure that you pad that as well.

If you have all this set up and have done your prep well you should be able to avoid getting in too deep.  It’s better to have the production plans worked up early than to suddenly discover, like the people in those Kickstarters that tried too much and found out that things were not ready when they needed to be.  Better to plan ahead.

The Let’s Build Series.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s