QMS Improvement - Where To Start?

[Randy]

Whether it's on time "delivery," "shipment," or whatever you want to call it is meaningless.

Really? When I'm told I will have something in my hands on X-day and I get it on X-day +4 it's not the same thing because it puts me behind with my shipping schedule to my client that needed your junk to finish what he bought from me to ship to him.

 

[Golfman25]

Really? When I'm told I will have something in my hands on X-day and I get it on X-day +4 it's not the same thing because it puts me behind with my shipping schedule to my client that needed your junk to finish what he bought from me to ship to him.

Yeah I know. But it is still semantics. We call it on time "delivery" but measure to shipment date. It still works because that is our target and, like I said, our actual legal delivery. Arguing about it worthless for the OP. He's go much bigger issues to tackle.

 

[Kurt Smith]

Ok, a lot of people work on semantics. Whether it's on time "delivery," "shipment," or whatever you want to call it is meaningless. In fact, for most shops like us (component contractors) legal "delivery" is at the shipping dock. Stay away from semantics. So you set a date to ship in your system and the question is whether you met that shipment date or not.

If you didn't meet that shipment date, then why not. That's where we do our pareto chart. We monitor things like capacity, incoming metal delivery, delays at plating/painting vendors, machine problems, tooling repairs, etc. Now our method of tracking these issues is really complex -- our production manager writes it down on his weekly shipment reports and we sit down and compile all the info monthly. Like any job shop, our biggest issue is capacity. So we look at that for ways to improve.

As for your part specific questions, do you have any ERP management system at all? Or just trying Excel and word? Have you done a map of your process to take an order thru production? Do you have any idea where the time sucks are? How do you get an order from PO to shipment? Good luck.

Those are exactly the kinds of steps I want to put into place. But ownership and other management here don't see the value in putting in the work to track information. They see it as waste of effort, and that there is not time for it when things need done quickly. I want to find ways to show them otherwise.

And no ERP system at all. We have an excel schedule of what is currently being worked on. But when jobs are completed the entries are deleted, and no unified record remains. We are extremely high mix low volume. So it all gets complicated in a hurry. Until now when jobs were completed the work travelers were filed away without ever being reviewed. I am planning to set up data entry and make a dashboard for keeping track of how we have done in the past.

What I want is to convince them that standardizing the work and applying data to our decision-making would be worth the hassle involved. I just don't know how.

 

[Bev D]

Yeah - creating a spreadsheet is not going to help much - the truth is hidden in the travelers. Your first act could be to understand if the delays are due more to quality problems or to simple flow / bottleneck problems. To do this you really need to get out on the floor. You might see this in a simple review of the travelers (looking for rejected parts and long wait times. But don’t caught up in analysis paralysis. This isn’t about precision and accuracy in the data (which isn’t going to be very precise in the first place). You only need to have a rough idea.

A stop watch is 10,000 times better than some spreadsheet with ‘remote’ data…I know. I’ve done this in large and small production lines. Buy a stopwatch.

Look for a book on mixed model flow design adn then get yourself into a good workshop. A simple google search should give you a few choices. My personal ‘go-to’ is from the Leonardo Group (I get nothing out this recommendation).

 

[mattador78]

We run our OTD or OTS to the day is completed and ready to be collected. We tell them what day it is going to be ready for collection or shipment. Unless transport is arranged through us, its then on them for when they get it. Worked well so far and no issues for audits as i show thats what we do, our only real lates are pro forma customers who unless they pay they aint getting it regardless of the promised date. I class it as its On Time for Delivery, as pointed out flat tyres, sinking ships and the effects of climate change are the courier company and the customers problem. If our drivers late with it then thats our problem but we have dedicated days and runs and so far having four licenced drivers for three trucks and only 2 on the road at one time the redundancy has worked well.

 

[Ed Panek]

"What we need is action on the front end as jobs come in. "

Try to involve the employees directly facing this requirement. I started out a long long time ago in production/service. I had hours upon hours to think about how broken things were. As we worked "and whined" wed have plenty of ideas. Dont let ideas go to waste. If the front line troops are involved they will be more eager to adopt any new processes you create. As the QM you just need to make sure it meets requirements, has been risk reviewed, and can be understood by top management. Set up a metric (Old way metrics vs New way metrics) That will be great to present at the Management Review. Inlcude metrics like "accuracy, Time, employee comfort, customer feedback (if any), Etc." If its radically different a dry run TPD might be needed to demonstrate results.

 

[John Predmore]

Another suggestion that might help you get started - as a small business, you may be able to tap into free outside help. At my startup firm, we learned there was a state grant to provide quality consulting services through a state-funded office. The university nearby was looking for student projects for senior engineering students. The local chamber of commerce has a quality staff that once a year partners their retired volunteers with small business organizations to provide problem-solving leadership and manpower. All that is required of the host company is a project charter, a liaison person and a subject-matter expert for a few days.

 

[Kurt Smith]

Yeah - creating a spreadsheet is not going to help much - the truth is hidden in the travelers. Your first act could be to understand if the delays are due more to quality problems or to simple flow / bottleneck problems. To do this you really need to get out on the floor. You might see this in a simple review of the travelers (looking for rejected parts and long wait times. But don’t caught up in analysis paralysis. This isn’t about precision and accuracy in the data (which isn’t going to be very precise in the first place). You only need to have a rough idea.

A stop watch is 10,000 times better than some spreadsheet with ‘remote’ data…I know. I’ve done this in large and small production lines. Buy a stopwatch.

Look for a book on mixed model flow design adn then get yourself into a good workshop. A simple google search should give you a few choices. My personal ‘go-to’ is from the Leonardo Group (I get nothing out this recommendation).

I agree that intentionally observed data and Gemba style decision making is much better than relying on spreadsheets. But a spreadsheet is still better than nothing.

I think I might have undersold both how little data we collect and how high mix our work is. We have hundreds of active part numbers moving through the shop each month. Orders can be for single digits or hundreds at a time. And parts range from a flat bracket that takes 2 minutes to pallet sized weldments that take days apiece. And in all of that, once the work is completed the documents are filed away and never looked at by anyone. The next time a part is ordered, if the details of how it was made before are not remembered by the operators we are basically starting from scratch. When RFQs come in, management has no way to measure how much bandwidth is left before new rush orders will cause delays.

Like a lot of businesses, this place started with a couple guys in a garage and grew from there. When they had problems, they relied on talented people and a flexible attitude to get through it. And that works great at the small scale. But the reason every successful large organization uses standards and consistent processes is because the larger you get, the more that reliance on talent and flexibility becomes a liability.

 

[Golfman25]

Those are exactly the kinds of steps I want to put into place. But ownership and other management here don't see the value in putting in the work to track information. They see it as waste of effort, and that there is not time for it when things need done quickly. I want to find ways to show them otherwise.

And no ERP system at all. We have an excel schedule of what is currently being worked on. But when jobs are completed the entries are deleted, and no unified record remains. We are extremely high mix low volume. So it all gets complicated in a hurry. Until now when jobs were completed the work travelers were filed away without ever being reviewed. I am planning to set up data entry and make a dashboard for keeping track of how we have done in the past.

What I want is to convince them that standardizing the work and applying data to our decision-making would be worth the hassle involved. I just don't know how.

The problem as I see it is you are going in with a preconceived notion of what you want to do -- setting up data entry and dashboards. Don't waste your time. Focus on what you have and work from there. Nobody wants the new guy to come in and start changing things. You say you measure on time delivery -- so that is your starting point. Only thing you need to add, is why the job was late. Start tracking that and when a trend appears, attack that problem. Then move to the next, etc.

First thing I would try would be to set up part files for repeat parts. That way everything can go into the file for easy review when the part comes back around. 30 years ago, we used to have stacks of prints, job cards and pricing cards all thru the office. Getting a new order was a hoot as they searched thru all that crap. New office manager came in and said "this is nuts." Created part files for all the parts -- in the file was the print, the job card, the pricing card, and any notes and other information about the part. Still have those files today. Whenever there is a question about a part, we can go to the file.

 

[Bev D]

All true and it’s why you’ll struggle to get leadership buy-in. They will dance with who brung them. They can’t understand what they’ve never experienced. I do think some looking at the travelers will be helpful but don’t navel gaze. And don’t discount that TPS won’t help with very high mix, low volume manufacturing…or you will fall into the same trap as yoru leadership. Someone who only seen white swans cannot beleive that there are black swans…