Message Thread:
learning scheduling
4/14/15
Well, I've been at the helm of our business coming on 7 years now. I took over just in time for the economy to tank which was actually a (painful) blessing. It taught me how to look hard at exactly how the business runs and keep it as lean and efficient as I could. Learn or die and we made it out alive.
Now we're thriving in the booming economy in our area. As we get increasingly busier, the most obvious thing that needs improving is my lack of scheduling prowess. I've never been good at it and the thought of even how to start figuring it out makes me run to do something else. However, it's starting to affect our business in an obvious way and I need to wrap my mind around it.
So, does anyone have any recommendations on "good reads" that can get me started down the right path? I know every company is different, but can anyone break down to basics the way they tackle scheduling multiple jobs?
Loaded question I know, but any help appreciated.
4/14/15 #2: learning scheduling ...
There are only two ways to do scheduling that I'm aware of: precision and gut feeling.
The problem with precision is that you have to have really, really accurate process times, or all the inaccuracies compound and you're left with a huge mess.
The problem with gut is that it depends on your experience. Sounds like you have some but not enough.
Hard as it may be, my only recommendation is to continue refining your gut by making mistakes and getting better every time. The key is figuring out why you were wrong so you don't make that mistake again. Also, really pay attention to what will need to be done, where things are at on a daily basis, and how far ahead/behind you are when you finish.
We have the "luxury" of mostly doing a lot of little jobs, so we start out scheduling kind of "loose". If all goes according to plan, then we can fit rush jobs into the holes or condense the schedule. Or, if jobs take longer we're already on schedule by simply not adding anything.
The problem with scheduling "loose" is that you have potential to lose a job if you quote too long a lead time. So you have to feel out your customer and make sure they can live with your time line. If they are not very responsive in telling me their needs I usually put out something like "Current standard lead time is X-Y weeks, but we may be able to produce in as little as Z if necessary.".
4/15/15 #3: learning scheduling ...
Website: http://optisol.biz
You asked, "Can anyone break down to basics the way they tackle scheduling multiple jobs?".
Scheduling multiple, concurrent jobs which share many common resources of limited capacity for improving on-time delivery, throughput and work in process is a challenging task in general. Most industries somehow manage it by an Excel application, a whiteboard or commonsense plus experience. I am not aware of any book or article that can give you practically effective guidance us in getting the best production schedules. Of course, the books like Eli Goldratt's "The Goal" may give excellent suggestions but it is not so easy to implement them on shop floor and take advantage of them. The devil lies in details mostly.
There are a few powerful, scientific, cost-effective software tools (like our Schedlyzer Lite) for scheduling multiple, concurrent jobs subject to finite capacity of resources. They provide reliable prediction of workflow, bottleneck formations and job completion times and facilitate fast and extensive what-if analysis and proactive capacity planning. However, their main drawback is that they need good information about operations of each job such as estimated hours, resource requirements and precedence relations, etc. In my experience, many woodworking people get disappointed with such data needs.
Multi-Project Scheduling
4/16/15 #4: learning scheduling ...
we are looking at crystalline for our countertop shop. It looks at production and installation and places a job on the calendar
Our casework shop uses first in first out basis on an excel spreadsheet. There really is no other way. Our columns are customer , job, job in, promise date, drawings submitted, drawings returned, materials ordered, materials recvd, sent to code, ready for assembly all conditionally formatted in red for no text. It changes color to green or clear when text is entered as each step is completed. Did it on google docs.
Seriously, first in first out has helped us a lot. If a job is not ready you can start the next, but as its drawing to a close, we need to go back to the one in line before it.
In another thread Tim Schultz and Pat Gilbert talk about what you put into something like software, can you get the ROI in time and $$$ ??? This really has to be thought out before buying the software Will you have time to use it. ? Will it require a body to feed it and how many hours a day ?
4/17/15 #5: learning scheduling ...
There are some tertiary benefits that are hard to quantify in ROI. Learning database changes the way you think about your business.
Answers to questions come from value lists so this makes you proactive consider not just what the questions are but what the allowable range of answers are. If you frame the questions properly the answers can produce a trigger to cause specific activities to happen. In this way database can not only leverage your brain power it can reduce the amount of time you have to spend in ad hoc response or mopping up from bad ad hoc response.
4/18/15 #6: learning scheduling ...
The more repetitive your product the easier it is to use scheduling software and perhaps the less relevant. Most is based on past timings, how much variance can there be in those? How much overlap? Can you always start step two when 50% of step one in complete? Depends on the size of the job.
At one point we bought a complex, software based on critical path method. It proved too time consuming for our product line, we went back to Excel. Using conveyors also made it easier to see if there were flow problems.
4/18/15 #7: learning scheduling ...
Larry,
The advantage of repetitive projects is reliable estimation of task durations. However, the scheduling function may still be relevant when there are several similar (not exactly identical) projects at different stages of progress competing for shared common resources.
I do not have much respect for any scheduling software that does not allow overlap of tasks in a production schedule.
The critical path method (CPM) developed in the late 1950s for scheduling a single project ignores resource requirements of tasks and any software tool based on this method usually poses a major difficulty in generating production schedules without resource overloading. Such tools force users to do resource leveling manually in a laborious manner in order to eliminate resource overloading because their automatic resource leveling is not dependable. For this reason, those tools keep schedulers quite busy with generating and revising resource-constrained production schedules.
Project scheduling tools do not deal with typical manufacturing issues like production lead times, throughput optimization, etc. They do not find optimal start times for projects based on limited resource capacities as shown in the web page cited below. Moreover, they do not have appropriate user interfaces for manufacturing environment.
Lean Production Schedule
4/19/15 #8: learning scheduling ...
"Production becomes lean as shown in Figure 2 below when setup times are reduced and work orders are released at right times." Reduced or eliminated setup times along with balanced loading will be the key to better flow, reduced WIP, more predictable delivery, etc. Much easier to achieve with a consistent product system. The more variation allowed outside of that system the more irregular the flow becomes, no matter what the release date. As a store fixture manufacturer, I'm well versed on variations that fall outside of our primary methods. Outsourcing some of them works by transferring the variations but with some loss of control. Perfect manufacturing is either very in-elastic or less efficient. If all resources are fully loaded and have optimized flow, perfection has been achieved, until something changes. BTW, CPM is not restricted to one job! I grew up in construction, crews and equipment are always shared among all jobs. Construction is just scattered site manufacturing with weather as another variable.
4/28/15 #9: learning scheduling ...
Thanks to all. We're highly custom, so no job is the same but there are similarities. The hardest part for me is the different production times of each of my employees. If it were me or my foreman building it'd take 3 days. The others, are more like a week plus. If I could only clone myself a few times...
Cabinetmaker, the column headings was along the lines of what I'm looking for along with the 1st in 1st out idea. We've always got multiple jobs going at once and projects get moved up or delayed so it seems like everything is always in flux. Truthfully, we might need to focus on just getting 1 job done at a time and making everyone wait in the line. Thank you.
4/28/15 #10: learning scheduling ...
Pdub,
you said, "The hardest part for me is the different production times of each of my employees".
The tool I mentioned earlier addresses this issue in terms of employee efficiency level.
You also said, "We've always got multiple jobs going at once and projects get moved up or delayed so it seems like everything is always in flux".
In my opinion, the same tool is ideal for such scenarios.
|