Having a quicker production line is pointless if your team will spend the next day culling defective parts.

Slowing down production to accommodate inspections isn’t ideal either. You’re building inventory, taking your employee off of their task and will increase turnaround time.

You want to identify what you should be checking for, when to inspect it and how to fit that into your line.

What you’ll learn:

• Find the Defects That Cause the Most Waste
• Inspect Parts Before More Work Is Added
• Make Optical Inspection Fit the Line
• Keep Good Parts Out of the Reject Bin
• Turn Inspection Results Into Process Fixes
• Measure Good Output, Not Just Speed
• Wrapping It Up

Find the Defects That Cause the Most Waste

Some defects are minor. Others lead to rework, scrap, or customer complaints. There may be instances where you need to repair, scrap or receive a claim from the customer due to defects.

It would be nice to check for all defects that can occur but this may not give you the answer you are looking for.

There are several things to look at when determining what to check for on your line. Look at parts you reject, re-work and claims you may have received. Once you have your list, determine:

• What defects are wasting the most time/material
• Which defects will most likely end up in the hands of the customer
• Where in the production process these occur
• What is currently being used to identify the defect

These will allow your quality team to begin evaluating which defects need to be focused on first. There are many methods to detect these, from checking for missing parts to blemishes that will not be noticed by the customer.

Inspect Parts Before More Work Is Added

When inspecting parts it’s best to do so as early as possible in the production process.

If there is damage to an edge of a part, you can remove it from production once identified. But if it’s not detected until it has gone through assembly and additional processes you are wasting time.

You want to have your inspection at or near where the defect is occurring. This will allow less time for it to pass through your other processes.

Visual defects and dimensions can be inspected using AOI Systems. These can be used to detect shapes, size and verify that parts exist on the part. There are many applications where an Automated Optical Inspection can be beneficial to your production line.

As mentioned earlier, the key word is Visual. If your camera cannot see the defect then there may be another process that you need to look into detecting the issue.

Try your method on a known good and bad part to see if you can detect the defect at that point in your production.

Make Optical Inspection Fit the Line

Just because you have an AOI System that can detect a defect, that doesn’t mean it can do so fast enough for your production line.

Determine how much time you have to inspect each part. Make sure you are testing your whole process, not just the image capture time. There may be multiple steps that your line will have to go through to inspect a single part.

There are many ways to reduce these times. Using a fixture that holds the part in place, finding a light that allows you to see your desired defect and inspecting multiple things from one position (if applicable to your part).

Depending on your application you may be able to run your inspection inline with your production line or have a dedicated station that has the ability to handle your throughput. There is no one way that is better than the other.

Many manufacturers are looking to invest in Vision Systems. According to Deloitte’s Smart Manufacturing Survey, out of 600 senior manufacturing executives at large manufacturers with US headquarters or operations, 28% prioritised vision systems in the next 24 months as one of their top 2 technologies. While vision systems can help you identify defects it doesn’t always translate to increasing line speed.

Make sure to test your full process at your current production speed. Including any change over and difficult to identify parts.

Keep Good Parts Out of the Reject Bin

There are two errors that you can get from an automated defect inspection system. You can get false rejects and missed defects. Both will have an impact on your business but false rejects can cause issues by having an overabundance of parts to manually review.

False rejects occur when a good part is detected as having a defect. This can occur when the part is not properly positioned, lights change or if there is an acceptable variation in finish that you didn’t know about.

By relaxing your inspection criteria you may reduce false rejects, but may not be identifying all of the defects you need to. Test your system on known good, bad and marginal parts and evaluate what it is rejecting that should pass and vice versa.

Allow your employees to review parts if necessary, and make sure you have a plan in place for those parts that may need additional review.

Turn Inspection Results Into Process Fixes

Knowing you have a defective part isn’t going to help if you don’t know why.

Using data from your inspection process can help you find what is causing the defect. If you can gather information about your defect, time, lot and the machine or process associated with the result. You can begin to see a trend of what may be causing your issue.

Perhaps you are detecting a blemish on a surface, it could be caused from the previous operation that ran. By looking into that tool/process you may be able to address the fault without slowing unrelated machines instead of slowing down all of your machines. You may not know for certain that it is coming from that process, but it will allow you to narrow down your search.

Another way to improve your processes is by control charts,, this can help you detect when something changes with your process and dig deeper into why it changed.

Allow your quality department to communicate with production when defects occur. Have them make a note of what they changed and continue to monitor if the defect decreased or went away completely.

Measure Good Output, Not Just Speed

After implementing changes to your line, are you producing more good parts per hour?

Here are some ways you can measure the success of your line.

• First pass yield – how many parts make it through with no repairs
• Good parts per hour – how many good parts leave your line per hour (subtracting defective parts)
• False Reject rate – How many parts you think had a defect but actually don’t
• Missed Defects – faulty parts that passed the inspection step and were detected later
• Time spent reviewing parts

Collect data from your line before you implement your changes and run a comparison after your trial period. You may want to look at similar parts that you are producing and see how they stack up. Take into consideration any production increases/decreases, manpower, etc.

As you can see there are trade offs that can occur when you increase the speed of your line. Are you missing more defects, but getting parts through faster? Maybe your line is catching more defects and your production is dropping. The goal is to produce more good parts with less repair work and quality issues downstream.

Wrapping It Up

You don’t have to sacrifice speed on your line to detect defects. Know what you should be looking for, when to check for it and how you can incorporate it into your line. Then continue to look for improvements to your processes and measure your good parts produced.