Having been in the industry making bone screws, I have found characteristics that proper control (including last off) will catch most rare conditions. It will not prevent them, but it will catch them. That will provide much more confidence for reducing inspection during final inspection - especially for template characteristics. As you know, the key difference between acceptance sampling and control is control reviews the part in time order and acceptance is random. Time order allows you to find a condition and remove any previous that show it back to its origin. So, if it is not statistically predicted, it will be found and contained.
The main exception is the chip impression. That may be able to be minimized with coolant flow and direction, but on many materials - and titanium has its issues - it is not possible to mistake proof. 100% visual is required - both at the machine and final.
But, there are dimensional characteristics that elude mistake proofing but can be better controlled - and small drilled holes are notorious example. You really do not know when a drill is going to fail - it may break or wear much differently than another drill. For that I inspected 100% with pin gages at 75% of the tolerance. They used to gage at 100% at full tolerance - and guess what happened? Yep, they would fail final. Goes back to what I always say - manufacturing does not own the full tolerance, if anyone does Final Inspection does, but manufacturing does not. It would not accommodate processes after the drilling - such as deburring - or even little variations like gage tolerance and fit. After changing the manufacturing floor gages to 75% - and keeping their mitts off of the gages in final inspection, final inspection was a breeze.
So, when final inspection is required the best approach is to make sure whatever you make will pass it. That cuts your overhead way down - you only inspect it once. It helps keep shipments on time. Mistake proofing is one good approach if you can do it. For swiss machines, most of the mistake proofing comes in clever programming and tool choice. I suppose dimension proofing, with durable controls on the shop floor, is another.
The main exception is the chip impression. That may be able to be minimized with coolant flow and direction, but on many materials - and titanium has its issues - it is not possible to mistake proof. 100% visual is required - both at the machine and final.
But, there are dimensional characteristics that elude mistake proofing but can be better controlled - and small drilled holes are notorious example. You really do not know when a drill is going to fail - it may break or wear much differently than another drill. For that I inspected 100% with pin gages at 75% of the tolerance. They used to gage at 100% at full tolerance - and guess what happened? Yep, they would fail final. Goes back to what I always say - manufacturing does not own the full tolerance, if anyone does Final Inspection does, but manufacturing does not. It would not accommodate processes after the drilling - such as deburring - or even little variations like gage tolerance and fit. After changing the manufacturing floor gages to 75% - and keeping their mitts off of the gages in final inspection, final inspection was a breeze.
So, when final inspection is required the best approach is to make sure whatever you make will pass it. That cuts your overhead way down - you only inspect it once. It helps keep shipments on time. Mistake proofing is one good approach if you can do it. For swiss machines, most of the mistake proofing comes in clever programming and tool choice. I suppose dimension proofing, with durable controls on the shop floor, is another.