Scheduling Blocks/Crew Size & Resources

We are currently trying to get our schedule over into Epicor. The issue I am running into at the moment is how am I to schedule multiple employees (welders and assemblers) to an operation. Should I be using multiple scheduling blocks or should we have a larger crew size? The other issues is that based on Time Off (out sick, PTO, etc.) I could have 2 employees on an operation one day and then 3 the next. Maybe even go down to one. I am not sure how to have our schedule reflect these changes. Should I always assume I am going to have one employee on an operation?

Also should we be setting up our employees as resources or should it just be a general “Employee”?

The way is was taught to me, the Operations are the Verbs and the Resources/groups are the Nouns.
Operations for Welding would be Welding
the resource/resource group is where you want your capacity to live.
With out knowing more about the assembly area its hard to say what the best set up is.

Well, IMO, those issues are not related to Epicor (or any scheduling engine on the market), they are “philosophical”.

In the end, it all comes down to how much energy/time you are willing to invest on the production schedule vs the precision of the schedule. It goes like this:

image

Here are some questions that you need to ask yourself first:

  • Are you in a “High Mix/Low Volume” environment, where each and every job, BOM and BOO are different ?
  • How accurate are your operations time estimates ?

For example, if your welding operation has a time estimate of 10h but in reality the real labor hours clocked on the job goes from 5h to 25h because of multiple factors (bad clocking form employee, each job is different, etc), the time/energy required to update daily the number of employee required on each operation of each job is a total waste of time. You would be at this point of the curve:

image

Meaning that you can spend 10 man/hour per day updating the schedule but your schedule will never be more precise or of “Better quality”.

If you are in this situation, you can consider the schedule as an “Objective”. Meening that you can schedule the production on a weekly basis to provide the production with a “realistic objective” and every day the production team tries to make the best decisions possible to meet the objective.

For example, for the issue of “what do I do if Jason is sick wednesday ?”, the answer can be nothing. The objective stay the same but you make decisions on the shop floor to make up for it.

On the other hand, if your time estimates are very precise and you are operating on an assembly line were each job has exactly the same BOM and BOO, then you may be at this point on the curve:

image

In this case, it may be beneficial to dedicate people to the daily update of the schedule since the gain of precision is great.

So, the answer is “It depends”.

Good luck

8 Likes

This is so accurate and very helpful. Probably a bit of “tough love” I needed to hear. I cannot micro manage this schedule on a daily basis. Our BOM and BOO and times on operations are different every job as we are not a high production shop. We are a custom fab shop.

Since we are a fab shop and times on the operations can vary job to job, would it be more realistic to schedule 1 resource (not a specific employee in Epicor terms) to each operation? If my leads are able to put more resources (physically and not in Epicor) then great we are earning our time quicker?

Also as I am thinking about it, what happens to the scheduled operation as we finish our time but we are not complete? Does it push out day by day until the operation is claimed complete? Using your example from above, the estimated time given is 10h but we run into whatever issue and we have used all 10h and we know we are going to need 25 more, when the scheduler is run will it keeping adding to the operation one day at a time?

Each Resource Group needs at least 1 resource.
The Operation needs either a Group or Resource on the Operation, but it can have more than 1 on the Operation. Think about who/what is doing the work. So if you have a work cell that has both Welders and Assemblers working in that same cell you could have it as 1 op with 2 resource groups, or 2 ops with 1 each.
An example of when i have had 1 resource group with more than 1 resource is in a machine shop with a press brake. we had 3 different size press brakes. so i had 1 group and 3 resources in that group.
The Operation was always PressBrk with the Group because it didnt matter with press brake used it. If i needed to use a larger press brake we would edit the Method of the particular part that had to use the larger one.
Are you running MRP with Finite capacity?

Remarkable post. You captured the essence of ERP scheduling in a single page.

Well done.

1 Like

The goal (and challenge) in your scenario is to generate realistic job lead times. If an operation is usually done 90% of the time by two employees, then configure it that way. Otherwise, your job lead time will be too far off. Keep in mind that purchasing of some components/raw materials are driven by the start date of operations when the MRP is run. Also, obviously, it will be impossible to predict the ship date without good lead times.

Operation labor time vary a lot from job to job ? You can add buffers to a certain degree to make sure that the lead generated is realistic.

If an operation takes longer than planned, yes it will push the next operation back day by day. If your operations takes waaaaaaayyyyyy longer than planned, it would be a good idea to update the estimated time of the operation. If you do not update the value, the scheduler does not know how many work hours remains and therefore cannot generate a realistic lead time.

1 Like