What's new
What's new

New Fabrication/Machine Building Shop Workflow Development

Mross506

Aluminum
Joined
Oct 6, 2019
Good morning Gents!
I am fairly new to this community and have only posted equipment related questions to this point. The quick summary is that I run an industrial maintenance team for a quickly growing manufacturing company that has a very proprietary process and prides itself on self-sufficiency. For the first couple years our maintenance techs also performed the secondary role of leading the new machine installs and taking on small machine builds. All of our fabrication work came from outside shops. After years of struggling with timeline and quality issues, we decided to bring 95% of our fab work in house and split the best techs off our maintenance team and make a full time "Machine Building" department. The growth has been fast enough to warrant a 250,000 sq ft building expansion in 2022 and again in 2023. Along with $15M in equipment that was a combination of OEM designs, collaborations, and 100% in house builds. Quickly our team grew into 6 full time Technicians that handle installation and commissioning, 5 full time welder/fabricators, and 1 painter.

Here is where I need help: I am a 20 year multi-craft maintenance electrician and while I have been blessed with a very successful career of building/overhauling maintenance departments, I have ZERO first hand experience working in Fab Shops. This makes it extremely challenging for me to be able to build systems/processes/procedures for a range of items. I have alot of good guys and we have been getting by for the last year or two but I don't want to just get by off hard work. I want to build a team that eventually is a fully functioning, world class machine builder. We are blessed to be able to hire top skill employees since we are not worried about a traditional overhead like a business would be. We are able to pay well and still save 40-50% on each machine by building it in house, while also working thru some really aggressive timelines. So with the guys that are making $30+ an hour, I should be able to build a shop structure that allows us to be efficient and repeatable regardless of the job.
  • Work Planning (What is the best practice once the set of prints for the assembly has been rec'd from engineering? Time estimates need established that aren't guesses, material lists developed, etc)
  • Work Scheduling (Currently all work is just given to the lead welder and he assigns it out as he sees fit. There is no formal tracking, monitoring, etc. I know we are losing time here!)
  • Welder Job Classes & Pay (Welder pay ranges from $18-$36 but we don't have clear defined job classes that specify the requirements for each job class)
  • Scheduling the assembly work (Do I need to hire 2 different job classes or can I create a hybrid that can do both and pay them well for it?)
    • Prima-Dona Welders (We struggle with some of the welders that just want to sit at the table with a stinger in hand and have someone else do all of the prep and post-weld clean up)
  • Quality Checks (How do you build the process? Is it the same for every component built?)
The list goes on and on! My hope is that within this awesome community there is a gentleman or two that would be willing to spend 20 minutes every couple weeks chatting with me via skype/teams in a mentorship role. I am sure this is not the standard request but I hope you will find my drive, passion, and appreciation of you time as being genuine. If there are resources, books, etc that you think would be beneficial, please share them and I will gladly devour them! I live an hour east of Columbus, Ohio if anyone happens to be in this area. I'd love to buy you lunch!

I can also break these questions down into individual forum posts if everyone thinks that would be worth the time in doing. I apologize for how green I am in this world. One thing the Army did was teach me how to percervere regardless of the situation. That single belief has opened the doors of opportunity for me more times than I can count. Each time I do my best to educate myself and hit the breach running!

Thanks for the help!
 
Hi, the following is briefly what I have understood from your post as a job shop scheduler. I may have a biased understanding of your requirements.

Your fabrication shop faces a challenge of efficiently making a range of items using very limited resources which include 6 full time Technicians, 5 full time welder/fabricators, and 1 painter.

The lead welder currently schedules your production using his knowledge, experience and commonsense but there is no formal tracking of work progress on shop floor. However, you are finding situations where some workers have to sit idle waiting for others to finish some work.

One of your requirements seems to be a good scheduling solution that helps with optimal production scheduling, tracking of work progress, prediction of work progress and product completion times and probably capacity planning.

I have been on this forum for many years but I rarely get involved in the discussions. If I have understood one of your requirements correctly, then we can have a brief chat over web to discuss further.
 
That is an accurate analysis of my situation. We are using a project management software that is sub-par for job scheduling but do not have the budget approvals to purchase an ERP system. (We also don't have the need for the accounting function of the system so investing in one becomes even less likely.)

Along with a basic question of managing the schedule, I would also like to learn more about calculating job time estimates when working in a world that doesn't see much in the way of repetitive work.

If you would have time for a quick chat, please message me and let me know how/when you are available to proceed.

Thank you!
 
That is an accurate analysis of my situation. We are using a project management software that is sub-par for job scheduling but do not have the budget approvals to purchase an ERP system. (We also don't have the need for the accounting function of the system so investing in one becomes even less likely.)

Along with a basic question of managing the schedule, I would also like to learn more about calculating job time estimates when working in a world that doesn't see much in the way of repetitive work.

If you would have time for a quick chat, please message me and let me know how/when you are available to proceed.

Thank you!

Hi Ross, Almost all project management software are not sufficiently powerful for scheduling complex production in job shops because they cannot automatically generate a production schedule without resource overloading.

In spite of many claims by ERP software vendors, I am still unaware of any ERP system that has a truly powerful module for scheduling production in a variety of job shops.

I sent you my contact phone number through a private message. Looking forward to talking to you over phone.
 
Are the machines you are building one offs or bespoke to each customer or can you program in some standard parts manufacture that will be required at some point and can be ready for when needed. we build machines here and find some of the order in parts like hydraulics and more so now electrical components can have ever increasing lead times and so difficult to get all parts to assembly at the right time, quite often having some parts made and sitting for months before other parts arrive to be assembled. with fabrication the right machinery can make a huge difference to the time to do the works, whether it's automatic cut off saws or CNC profile cutters and CNC benders it can reduce labour costs and increase quality however it's not always the case. We have a multi-skilled workforce that can diversify to the needs of the shop and no doubt some are better at particular tasks than others so it's a case of planning the works to get the right guys on the right jobs at the right times. this is where having standard parts list to keep guys busy working makes parts when you need a buffer to reorganise the guys priorities.
Also if you have ten or so skilled guys producing goods you can justify employing a quality manager to ensure everyone's work is up to scratch, all the materials are as ordered and stored correctly, cutting lists are made up to utilize stock efficiently, any welding or consumable or training certificates are kept in order. it's sometime the little things that can add in unwanted time delays or even worse poor quality work being produced. this can give you the breathing space to stand back and see where the bottlenecks are in the process.
 
We are the customer in this case so we have a decent amount of ability to standardize to certain components across multiple builds. (You hit it on the head with lead times!) We try to stock a base level of the most common parts but have a large variety across what we are building so the only real opportunity to truly standardize comes in the selection of safety fence interlocks, light curtains, FRL units, etc. As of now we have 5 full time fabricators and 6 full time installers. I would expect that to double in the next 2 years as we bring more and more work in house.

I am definitely missing the full time scheduler/quality inspector. Right now I am spending a lot of my time trying to build out the system, create procedures, etc in order to then have the team just need to follow them but even this is a struggle because of my lack of fabrication knowledge and how fast our shop operates. It doesn't make sense to spend more time talking about a plan than just executing but at some point I know we cannot sustain that as the shop grows...I also have the tendency to want to fix everything all at one time or struggle with rolling out one piece of the puzzle when that piece doesn't work well because a couple of its key components are yet to be standardized....

Our Eng team is rolling out a massive overhaul to our CAD format that puts alot more of the part info at our finger tips so that it is easier to kit parts...It is slowly coming together, just really cumbersome and hard to keep momentum.
 
I want to build a team that eventually is a fully functioning, world class machine builder. We are blessed to be able to hire top skill employees since we are not worried about a traditional overhead like a business would be. We are able to pay well and still save 40-50% on each machine by building it in house, while also working thru some really aggressive timelines. So with the guys that are making $30+ an hour, I should be able to build a shop structure that allows us to be efficient and repeatable regardless of the job.
  • Work Planning (What is the best practice once the set of prints for the assembly has been rec'd from engineering? Time estimates need established that aren't guesses, material lists developed, etc)
  • Scheduling the assembly work (Do I need to hire 2 different job classes or can I create a hybrid that can do both and pay them well for it?)
    • Prima-Dona Welders (We struggle with some of the welders that just want to sit at the table with a stinger in hand and have someone else do all of the prep and post-weld clean up)
My previous employer was a builder of custom machinery. One of the things they did was to make the engineers do more work themselves. A lot of companies have engineering design the thing, then throw it over the wall to production, and then maintenance. In the entire company, we had about 3 mechanical assembly techs. As an engineer, I was responsible for designing my part of the machine, then the parts were made either in-house or at a vendor. Once the parts came back, the engineers assembled the machine, tested it, took it back apart, installed it at the customer, and flew out to do PPM trips. If we were doing serial production, it wouldn't be economical at all, but it did produce very good machinery.

With the normal system, engineers do stupid things that make it more difficult to build and maintain. Most of the time, they never even learn that they should have done it differently. Getting the engineers involved more with building and maintaining will help them build better machinery.

I think there's a happy medium, and my old employer would have been better off with more techs, but the general concept was great.
 
This is much aligned with how we operate. Originally the ENG's performed a large portion of the work but as we grew it becasme more difficult to find enough ENG's to keep up with the design load. (Especially ones with hands on experience) The way we are set up currently. the ENG still owns the overall machine design & installation responsibilities but is really hands off once the print packet it created until the machine is ready for commissioning unless there are questions or changes required.
 
it's a case of planning the works to get the right guys on the right jobs at the right times

Utilization of multi-skilled workers surely provides a lot of flexibility in assigning skilled workers to operations in real time but generating a production schedule with optimal utilization of them for better production KPIs is a challenge for many people on shop floor. However, good scheduling software can help do it quickly and automatically.
 
I am definitely missing the full time scheduler/quality inspector.
You see this requirement probably because you are finding it difficult to manage the scheduling function in your shop right now. In my opinion, for a small shop like yours, creation and revision of sensible production schedules should not require more than an hour or two per day if a right tool is used. In this approach, most of the scheduler's time is spent on updating job status information everyday. Creation or revision of a right schedule may take only a fraction of a second even for 1,000 operations. However, there is no need to reschedule job shop production several times in a day.
 
This is much aligned with how we operate. Originally the ENG's performed a large portion of the work but as we grew it becasme more difficult to find enough ENG's to keep up with the design load. (Especially ones with hands on experience) The way we are set up currently. the ENG still owns the overall machine design & installation responsibilities but is really hands off once the print packet it created until the machine is ready for commissioning unless there are questions or changes required.
Unfortunately, a lot of where the engineers should get their experience is in between the print packet and commissioning. One advantage of engineer-heavy production is that you can have more new engineers without overloading your more experienced guys. A good portion of their time during the first two years will be doing non-design work, so that minimizes the amount of time the more senior engineers need to spend supervising newbies trying to design things. Also, make sure that you have at least one welder and one machinist who are willing/able to talk the new engineers through the manufacturing processes. It's far better to spend 15 minutes going over your draft with someone who knows what they're doing than it is to try to fix the problem later.

Where are you located?
 
You see this requirement probably because you are finding it difficult to manage the scheduling function in your shop right now. In my opinion, for a small shop like yours, creation and revision of sensible production schedules should not require more than an hour or two per day if a right tool is used. In this approach, most of the scheduler's time is spent on updating job status information everyday. Creation or revision of a right schedule may take only a fraction of a second even for 1,000 operations. However, there is no need to reschedule job shop production several times in a day.
I agree with that. Updating the job status', kitting assembly work, analyzing multiple main assemblies and all of their sub-assemblies to insure all the right pieces are moving thru the shop at the right time...The only time schedules change is when our estimated task time was not accurate or emergency jobs jump into the schedule. (Which we try to limit but still have fairly regularly)
 
The only time schedules change is when our estimated task time was not accurate or emergency jobs jump into the schedule.

The reasons for deviation of work flow from the schedule also include uncontrollable natural variation in individual process times, rework, unexpected interruptions in resource availability, changes in job priorities, etc.
 
My 0.03 € ..
You don´t require any more erp, jobshop sw, etc, mostly.
And none that currently exist would really help You - as-is.

You are in a dynamic environment making high-value items with complex dependencies where one part needs another -- and the lead time may be prohibitive.
And the lead times and deliveries and needs change in a complex dynamic manner in unpredictable ways, often within a single day.

Like when a super customer needs to bump something up the line, or a sub cannot deliver, or certain materials are delayed, endless issues.

What you actually really need is 2 secretaries, part-time preferably, inputting all the data available into a good scheduling platform.
Minimum 2, because your results cannot depend on a sec. being available, due to sick whatever.
Monday is a good alternative.

And secs. because neither your engineers nor your fabs can approach any of their subtasks calmly, rationally, with numbers.
? How many hours was this supposed to take ? 300 ?
... It´s been 600 hrs so far. Is there a supplier not good ? Can we improve our internal planning ?
And the secs cost 5x less.

You NEED numbers. Your engineers and fabs will both hate it the first 6 weeks, and LOVE IT after 10 weeks.

Your engineers will be much less pressed,
your fabs will be much more productive and relaxed,
and the secs will be there not to supervise them because they are not supervisors but attendant for both,
and will be asking "can I get You something to make it easier ..?".

YOU cannot do this personally.
It takes superhuman patience and people skills under tremendous pressure and endless hassle -- because it´s a real-time job with endless ring-zero interrupts for totally irrelevant silly things.
No sw can do this because the jobs and suppliers and parts and transport and issues are not linked 24x7 to IT or yourselves.

I´ve done this 3 times in 2 countries in 2 languages with 11x the results of previous operators.
With at least 4 different products, maybe 6.

You may not get the approval for this approach in terms of costing.
Probably wont.
But it´s extremely efficient and would pay it´s way in many multiples first year.

Suggest asking for 2 technical secretaries (semi-retired, very experienced, may be good).
And 150-250k for general setup and consultant on-site 2 months and offline 12 months.

Example:
All my HAAS sales guys hated my/our professional structured approach first 2 months.
All made at least double the sales of their best year ever, after it.
All loved it.
All pushed all the silly details onto the sales secretaries, and got on with their real job -sales.

Same applied to engineers.
They quickly LOVED not needing to worry about anything because c***a would take care of everything, including complex or expensive or heavy tooling transport, security, planes, car, hotel, whatever.

What You really need is a good kanban type board up all day - where You yourself can make decisions based on real data and not be running after emergencies everywhere.
You need to *look at* the kanban type data board, not trace data for inputs.

My approach is really cheap, and really effective, and You are guaranteed results.
Without someone inputting the data, all the time, You will never have a managed business.
It might be successfull, perhaps even very successful - but this will depend on You being extremely good, efficient, stressed, multitasked, and after everything all the time.

I´m suggesting what will likely work - and what it will cost.
Trying to do it for less -- is unlikely to work as well.

Hiring any of the pro companies for same will take 4x longer, cost 5x more, and deliver 2x-3x less results for 10x+ unhappiness.
 
You don´t require any more erp, jobshop sw, etc, mostly.
And none that currently exist would really help You - as-is.

You are in a dynamic environment making high-value items with complex dependencies where one part needs another -- and the lead time may be prohibitive.
And the lead times and deliveries and needs change in a complex dynamic manner in unpredictable ways, often within a single day.

Like when a super customer needs to bump something up the line, or a sub cannot deliver, or certain materials are delayed, endless issues.

What you actually really need is 2 secretaries, part-time preferably, inputting all the data available into a good scheduling platform.
Minimum 2, because your results cannot depend on a sec. being available, due to sick whatever.

No sw can do this because the jobs and suppliers and parts and transport and issues are not linked 24x7 to IT or yourselves.

You may not get the approval for this approach in terms of costing.
Probably wont. But it´s extremely efficient and would pay it´s way in many multiples first year.

Suggest asking for 2 technical secretaries (semi-retired, very experienced, may be good). And 150-250k for general setup and consultant on-site 2 months and offline 12 months.

My approach is really cheap, and really effective, and You are guaranteed results.
Without someone inputting the data, all the time, You will never have a managed business.
It might be successfull, perhaps even very successful - but this will depend on You being extremely good, efficient, stressed, multitasked, and after everything all the time.

I´m suggesting what will likely work - and what it will cost.
Trying to do it for less -- is unlikely to work as well.

Hiring any of the pro companies for same will take 4x longer, cost 5x more, and deliver 2x-3x less results for 10x+ unhappiness.

In the above reference, I skipped some sentences for focusing on the remaining.

I agree with your understanding of the complexity of job shop production under discussion. Production control and management become difficult to due to factors like material delays, engineering delays, unexpected interruptions in resource availability, poor estimates of task durations, changes in job priorities, acceptance of rush orders, rework/rejections, etc.

If the production system is highly chaotic due to these factors, planning and scheduling will be ineffective and the system becomes inefficient. Only continuous firefighting in real time is an effective solution in such case. However, if the system is not so chaotic, truly powerful scheduling software can help with efficient production control and management using few manhours. Such software can drastically reduce the cost and man hours required to implement your suggestion involving "2 technical secretaries, 150-250k for general setup and consultant on-site 2 months and offline 12 months". Some powerful software for scheduling complex job shop production may not cost more than $4,000 totally. Just one person with industrial engineering or operations management background can do wonders in this case using good scheduling software.

You said, "My approach is really cheap, and really effective, and You are guaranteed results.". I am not sure this approach is going to be relatively cheap, effective and guaranteed to be successful with salaries for 2 additional employees and a lot of consulting charge. In my opinion, what is really guaranteed here is the high cost with some risk of failure.
 
I am not able to see simplicity in the proposed expensive approach. You and the two secretaries might be just trained on how to do firefighting on a daily basis.
It is definitely not cheap at a $250k plus anoter $100k in salaries. I am curious what data is being tracked and why not just use an ERP system unless the systems just prove to unwieldy. We operate now off of a task level gantt chart that is ok but I still struggle in getting the Supervisor/Leads to operate off of it. We always revert back to just stacking tasks one after the other and whenever they are complete is fine...
 








 
Back
Top