An aspiring engineer learns a lot in school, but a lot more must be learned on the job. Much has been written about how that first job teaches you to work as part of a design team, as well as with different groups with a legitimate interest in the product (procurement, QA, and even marketing). We each recall different lessons from that first real engineering experience, in addition to becoming a better designer.
A friend and I shared some of the standout points about these lessons. They made the difference between being an engineer who merely knows the principles (and has built some actual projects) and an engineer who has been involved in product design, debug, and release to manufacturing.
First, there's the need for a real bill of materials, getting as many parts from the approved vendor list as possible. These components are already in use in other products, so they have already been evaluated and approved in terms of vendor credibility, ability to deliver, performance to spec, pricing, and other factors. Their subtleties and idiosyncrasies are also better understood.
Tolerances and variability
You can build one of something and tweak it so that it works, but what is the effect of component tolerances when it goes into volume production? As part of the design review, a design team must look at the typical and even worst-case buildup of variations in parameters ranging from supply voltages to passive component tolerance, as well as mechanical variability in factors such as component size and physical placement. It's no fun when units in the pilot run won't meet performance specs or when they can't even be assembled properly.
These drawings and notes show all the parts on the PCB, how they are placed, and in what order — something especially needed if there are unusual parts (brackets, clips, special connectors) in addition to standard surface-mount or through-hole ICs and components. There are also assembly drawings for the entire product, showing how the PC board, enclosure, and any other subassemblies go together — and in what sequence. It's no different from having the detailed drawing of a single part to be machined. Figuring out how to complete the various machining steps (and in what order) is often a challenge comparable to the actual machining.
Fixtures and jigs
These speed assembly and ensure it is done properly. This may be needed for parts that must be aligned (electro-optical ones are a typical challenge), assemblies that have to be placed just right, and torques and tensions that have to be set correctly. Of course, there are also test setups with special probe assemblies and configurations very different from those used on the prototype. This allows units to be evaluated quickly for basic functionality and calibrated if needed, and key specs can be verified before shipment.
All designs involve balance among performance, power, and cost, while keeping priorities in mind; there's no news there. But when you are looking at going into a production mode, there are dimensions to the compromise puzzle that simply don't exist when you just have to deliver a few working prototypes.
Finally, there's the working-with-others aspect. That's a huge story in itself.
These are the lessons that stood out for us. What design and nondesign lessons did you learn when you went from just designing a project to bringing it into volume production?