Scrap vs. Rework Classification - Scraping the resistor or reworking the PCB board?

[ACIQualityGuy]

I've been tasked with setting up a way to record, monitor and intemperate how scrap and rework effect our company. I've started in one of our more complex departments asking production to log relevant characteristics and have gotten some good preliminary data such as scrap and rework costs/percentages as well as a list of all the failure modes and what % of the total each mode is responsible for.

My ultimate goal is to work with our IT department to create a VB program that can be used to log this data so that it can be automatically converted into usable information/daily reports but I need to understand the process better before going this route. I see the value in capturing this information but have had a tough time determining if something is being scrapped or reworked.

Example:
If a resistor on a PCB board is bad and you fix it, are you scraping the resistor or reworking the PCB board?

At this point I've called this a rework and have found that nearly all of our products that don't pass our initial testing end up being reworked. In fact the only time I've seen a "scrap" is when someone has spent too much time trying to fix a product and decide it's no longer worth it.

How would you record the information given the example I choose?

Thanks in advance for the help!

 

[dknox4]

Re: Scrap Vs. Rework Classification

I've been tasked with setting up a way to record, monitor and intemperate how scrap and rework effect our company. I've started in one of our more complex departments asking production to log relevant characteristics and have gotten some good preliminary data such as scrap and rework costs/percentages as well as a list of all the failure modes and what % of the total each mode is responsible for.

My ultimate goal is to work with our IT department to create a VB program that can be used to log this data so that it can be automatically converted into usable information/daily reports but I need to understand the process better before going this route. I see the value in capturing this information but have had a tough time determining if something is being scrapped or reworked.

Example:
If a resistor on a PCB board is bad and you fix it, are you scraping the resistor or reworking the PCB board?

At this point I've called this a rework and have found that nearly all of our products that don't pass our initial testing end up being reworked. In fact the only time I've seen a "scrap" is when someone has spent too much time trying to fix a product and decide it's no longer worth it.

How would you record the information given the example I choose?

Thanks in advance for the help!

I would record the resistor as scrap, and the PCB as reworked.

 

[ACIQualityGuy]

Re: Scrap Vs. Rework Classification

^So you would record the cost of the resistor and the time to remove the old part as a scrap cost while the time to replace the resistor as a rework cost?

What media is everyone else using to capture this type of data (manually logging in spreadsheets, VB Programs, etc.)?

 

[Jim Wynne]

Re: Scrap Vs. Rework Classification

I've been tasked with setting up a way to record, monitor and intemperate how scrap and rework effect our company. I've started in one of our more complex departments asking production to log relevant characteristics and have gotten some good preliminary data such as scrap and rework costs/percentages as well as a list of all the failure modes and what % of the total each mode is responsible for.

My ultimate goal is to work with our IT department to create a VB program that can be used to log this data so that it can be automatically converted into usable information/daily reports but I need to understand the process better before going this route. I see the value in capturing this information but have had a tough time determining if something is being scrapped or reworked.

Example:
If a resistor on a PCB board is bad and you fix it, are you scraping the resistor or reworking the PCB board?

At this point I've called this a rework and have found that nearly all of our products that don't pass our initial testing end up being reworked. In fact the only time I've seen a "scrap" is when someone has spent too much time trying to fix a product and decide it's no longer worth it.

How would you record the information given the example I choose?

Thanks in advance for the help!

Your categories should be of causes and not necessarily things. Thus if you continually see rework because of "bad" components, cold solder joints, backwards diodes or ESD issues, you know what to work on. Don't worry so much about whether a thing has to be scrapped or not; it's costs that count, and excessive costs are related to causes.

 

[dknox4]

Re: Scrap Vs. Rework Classification

^So you would record the cost of the resistor and the time to remove the old part as a scrap cost while the time to replace the resistor as a rework cost?

What media is everyone else using to capture this type of data (manually logging in spreadsheets, VB Programs, etc.)?

I meant that the resister would be scrapped and the board reworked. If you are going to try to get and exact accounting of your scrap by each small component and rework by the minute, I predict you will have a new full-time job. I am in agreement with Jim in his post. You should be interested in how many boards you reworked or scrapped for what reasons.

 

[ACIQualityGuy]

Re: Scrap Vs. Rework Classification

Your categories should be of causes and not necessarily things. Thus if you continually see rework because of "bad" components, cold solder joints, backwards diodes or ESD issues, you know what to work on. Don't worry so much about whether a thing has to be scrapped or not; it's costs that count, and excessive costs are related to causes.

It sounds like you are referring to capturing first pass yield (and capturing failure modes as well as throughput #'s) rather then sorting between rework and scrap? I suppose as long as you are able to address the cause it will reduce both scrap and rework costs, it just doesn't identify which is the bigger problem, scarp or rework... Am I understanding you correctly?

Thanks for the help guys, I'm pretty green under the thumb and tend to think a little too detailed sometimes...

 

[JaxQC]

Re: Scrap Vs. Rework Classification

FTQ – First Time Quality looks at what ACIQualityGuy is talking about. It doesn’t care why it didn’t get to the next station (re-work or scrap) just that it didn’t the first time.

Cost wise you would have the time to re-work and the scrap material cost of a new resistor to add in but it is secondary ie.. result of a poor FTQ. I say secondary because it’s not really what you want to know (although what is takes to correct a problem vs the dollars cost does factor into the equation and perhaps to identify what area is bleeding worse to know where to focus). Knowing the dollar amount doesn’t do anything to change the problem. Identifying the drivers of the FTQ issues like Wynne points out starts on the path to know what to change or address.

Looking at FTQ is also an on-going item. Track it long term and when a factor changes (bad lot, machine off etc) it will jump out and be identified early. 500 bad resistors @ a penny jumps out as a bad change compared to 1 board @ 5 dollars. One board might not be worth investigating but the 500 qty means something is not going well & needs investigation.

 

[Jim Wynne]

Re: Scrap Vs. Rework Classification

It sounds like you are referring to capturing first pass yield (and capturing failure modes as well as throughput #'s) rather then sorting between rework and scrap? I suppose as long as you are able to address the cause it will reduce both scrap and rework costs, it just doesn't identify which is the bigger problem, scarp or rework... Am I understanding you correctly?

Thanks for the help guys, I'm pretty green under the thumb and tend to think a little too detailed sometimes...

No, not first-pass yield, or throughput. It's a question of categories and cost. Scrapping a three-cent resistor before it's soldered isn't the same as scrapping a $300 assembly, so they shouldn't be put in the same category. You need to find out how money is being wasted, and understand that "waste" implies an undesirable condition that can be responsibly prevented. Everyone knows that scrap is bad, but there can be levels of scrap that are economically acceptable.

Of course, to proper categorization requires data collection, and that's where you're hung up. It doesn't matter whether a given board is scrapped or reworked (per se); what matters is what it costs. How do you know whether to scrap a board or rework it? If rework is caused by a bad component, what's the cost of screening components compared to the cost of rework? You've been asked to find out where the money's going, even if the task wasn't communicated in those terms. You already know that some things are scrapped and some are reworked. Before you can do anything that has a salubrious effect on the observed phenomena, you have to know the costs, otherwise you'll wind up fixing things that don't need to be fixed.

 

[Bev D]

Re: Scrap Vs. Rework Classification

Jim has good advice.

let me take a stab: ask yourself if the knowedge of which is bigger (scrap or rework) is actionable? How does knowing that help you fix it?

The person tasked with fixing it is first going to ask: what is causing the scrap/rework? then you're right back to the exact failure mode that resutled in either the scrap or the rework...

always ask - how will this information be helpful and what will do with it, before decidign on your metric.

In my organization, we track by failure mode. we do track by occurence and total cost of fixing the failure (any scrap, rework and retest time, and any additional consumable costs) as ways of prioritizing what failures we go after first and monitoring progress.

 

[ACIQualityGuy]

Re: Scrap vs. Rework Classification - Scraping the resistor or reworking the PCB boar

Fist off, thanks to everyone who has responded, especially so quickly!

I think it's getting more and more clear that I was over-complicating things a bit by trying to separate scrap vs. rework. Instead, based on what everyone else has said here, I will focus on identifying a couple of key metrics:

1. Specific failure modes/frequency of each
2. Time spent/cost per failure mode
3. Overall cost/time of "failed" items per department/Company wide

By being able to identify what the biggest issues are (via #2 in terms of cost/time) and being able to identify the specific failure mode associated with that cost (via # 1) and the extent to which each failure mode has an impact on the total (extended cost), I think I will have actionable information that will aloow me to focus on what's most important. I also think it's a good idea to establish a baseline per department/company wide to monitor our progress which is why I choose # 3...

As Jim basically stated, I could then take this information gathered vs. the cost to implement a solution to decide if the problem merits action.

Any other thoughts? Am I still spinning in circles or does this make some sense?

Thanks again for helping me get on the right path