Wow! I think you'll learn so much from this internship experience and, if you do this well, will have a crapload of stuff to put on your resume.
I think your mode of attack is great and the additional suggestions are spot on. I work in quality control. We're not ISO. These are just some ideas. Use them, dont use them.
- Find out what the complaints typically are from customers (whichever way is kosher for you to do so, internally from your current company or from the customers themselves). Depending on who you talk to, you might get things like "Oh, its the wrong color" or "the memory isnt fast enough" - Those are moreso design issues. Its nothing you can prevent from going out the door because its ALL going out like that. I find you get this from the sales guys and some of your customers.
- Something to keep in mind: its obvious on paper but sometimes in practice we forget this . . . . You cannot inspect quality into a product. If something is recurring, talk with the production managers to see if its something that is inherent in the product that you'll have to just deal with (design flaw, might have to do with one of your suppliers, who knows? Yall can tackle that if you want but I personally would stick to setting up some QC checks) ... or if its something that is inherent in the process that might can be fixed. Identify what is ready to ship, what products could be reworked (if any), and what would make a product flat out rejected. So, at our company, each batch has thousands of units. We ready to ship that sucker if there's less than 1.0% defect. We dont check every unit, we use a sampling plan. We don't check for every possible thing that could go wrong; we check for characteristics that do go wrong. We dont do a sampling plan for packaging; the packagers are pulling from the same boxes so we spot check it every now and then. Someone else verifies the artwork is correct; quality is everyone's responsibility. We check one unit every now and then to make sure the final product looks right. Everyone knows what to look for. The parameters of pass/fail are clearly defined. Use pictures. Whatever it takes. Most importantly, our process CAN put out less than 1% defects. If your process, on average, puts out 2% defects but your spec (or more importantly the customer) says no more than 1% defect ... management needs to be financially prepared for that because 2% defective material is what your process puts out. This is where you can look into control charts and whatever to re-evaluate your specifications (are they realistic?). You said you dont have past data but you need the data to do this. And, if the first step is acquiring the data ... start acquiring. But, again, only if its necessary.
- What is your current process? Ask the supervisor or the lead training guy what the process is. Then watch, observe. Are the people ACTUALLY doing it? Dont comment on if things are good/bad (ie, judge). What are they currently doing? I also ask the people on the line (if they've been there awhile) what are some of the issues they see. They are honestly there every day and have a different insight. Depending on who you ask, might be relevant/might not. But again, see what your scope is in terms of how far you can stick your nose in.
- Identify what needs to be checked, quality wise. Easier said than done. Also, quality checks dont have to be done by a QC department. I work at a tableting company and production pulls tablets off every so and so minutes to check weight, thickness to make sure things are in spec. They write it on a form. At any time we have the right to verify. But we trust they do it correctly. Do you need to write procedures? What is a pass/fail? Do you need to make forms? Who checks it? How often?
- Would you do an in-line check to prevent the issue from getting too far into the process? Or do one final check at the end? And when problems arise, does it get rejected immediately or is it fixable (ie, you can rework/repair it)? Or does it get shipped out anyways? Do you need a sampling plan or is this something like a car where every single unit needs to be checked? Do you damage product you need to inspect by opening it up or breaking seals to get inside; if possible, check it earlier (in-line).
- Safety first. Hardhats, PPE, etc.
- How do you denote something is "ready to ship"? The person checking it knows but how does the dude driving the forklift know its okay to put on a truck?
- What is management willing to provide in terms of funding and long term support for a quality system? If you set up something it might slow the process down, inspection takes time/people, rejecting materials (in the short run) is immediately seen as a loss money-wise. . . . Some new kid telling people they cant ship something out that they've shipped out previously can really irk people if it has been the status quo for a long time. You NEED support of management. I dont know what your company's culture is like!
- Are all the checks that you implemented necessary in the first place? Also, if you have suggestions to make the process "better" - sometimes new changes make new problems. But, on the flipside, sometimes people have implemented something that isnt necessary to the process and is, quite frankly, a waste of time. So asking, "What is the purpose of this? Do we need to do it?" might not be a bad idea either. Always keep that in the back of your mind.
If you do write down production's modus operandi, I personally would do something quick and dirty and then focus on the quality aspects. If you have time at the end, then go back to the production manual. We've passed quite a few audits with "on the job training" as an answer but since we dont have high rate of turnover and we're a small company ... it suffices. A general process flow chart should start the company off in the right direction. Come back to some type of production training manual after you have a strong idea of the process and figure out all the buzz words.
Sorry, I got chatty and wrote this fast. And Im learning this too so take this with a grain of salt!