I agree pretty much with you - hard to be sure as I get dizzy reading your emails that don;t have an occasional paragraph break :)
Bottom line: Keep your capacity modelling as simple as possible to achieve good enough realistic modelling of your environment.
No model is perfect as it doesn't know "Joe's wife is having an emregency apendectomy next month & certain set ups will be not done until he returns", or "that 20 year old NC machine will be down for two weeks", etc.,... Few actually adjust their models to match real world daily changes to it as they occur.
Where Epicor really goofed is poorly explaining (even to there consultants) that resource groups & resources are NOT really they equivalent to the much more common "workcenter/department" paradigm that most finite scheduling packages use. They are simply the exposed building blocks to produce something similar to what 98% of the systems out there do.
It is really the Standard OPerations that are more intuitively similar to Work Centers.
By creating the basic contraints of capacity at the resource/resource group level, it is the combination of these set up in the Standard OP that gives equivalent capacity modelling behavior similar to what most have called Work Centers during their careers.
I think Epicor finally is starting to 'get it' with the addition of Capabilities. But even with this they are using lanuguage conventions few are used to and it is doomed to cause confusion. (I've yet to meet a consultant who truly understood it well enough to explain it clearly).
Also, since the Global itself is laced with bugs (two major ones fixed in 405 & the final one impacting us to be fixed in 408), it is all moot anyway. You're still going to get flawed schedules & since Vantage is backward scheduling biased (not even supporting a fixed start date constraint for forward scheduling... Basic 1970's MRP II capability), it almost gaurantees the resulting schedules aren't realistic.
Honestly, if your environment is that simple, why not go KanBan and eliminate the overhead entirely?
Lean is where it's at (and maybe Epicor will get that some day.)
Rob (Brown)
Bottom line: Keep your capacity modelling as simple as possible to achieve good enough realistic modelling of your environment.
No model is perfect as it doesn't know "Joe's wife is having an emregency apendectomy next month & certain set ups will be not done until he returns", or "that 20 year old NC machine will be down for two weeks", etc.,... Few actually adjust their models to match real world daily changes to it as they occur.
Where Epicor really goofed is poorly explaining (even to there consultants) that resource groups & resources are NOT really they equivalent to the much more common "workcenter/department" paradigm that most finite scheduling packages use. They are simply the exposed building blocks to produce something similar to what 98% of the systems out there do.
It is really the Standard OPerations that are more intuitively similar to Work Centers.
By creating the basic contraints of capacity at the resource/resource group level, it is the combination of these set up in the Standard OP that gives equivalent capacity modelling behavior similar to what most have called Work Centers during their careers.
I think Epicor finally is starting to 'get it' with the addition of Capabilities. But even with this they are using lanuguage conventions few are used to and it is doomed to cause confusion. (I've yet to meet a consultant who truly understood it well enough to explain it clearly).
Also, since the Global itself is laced with bugs (two major ones fixed in 405 & the final one impacting us to be fixed in 408), it is all moot anyway. You're still going to get flawed schedules & since Vantage is backward scheduling biased (not even supporting a fixed start date constraint for forward scheduling... Basic 1970's MRP II capability), it almost gaurantees the resulting schedules aren't realistic.
Honestly, if your environment is that simple, why not go KanBan and eliminate the overhead entirely?
Lean is where it's at (and maybe Epicor will get that some day.)
Rob (Brown)
--- On Thu, 3/5/09, rob.bucek <rob.bucek@...> wrote:
From: rob.bucek <rob.bucek@...>
Subject: [Vantage] Re: Repost....Resource Group configuration....HELP
To: vantage@yahoogroups.com
Date: Thursday, March 5, 2009, 5:16 PM
Ordinarily i feel a strong impulse to outright disagree with any advice Epicor may dispense out where sceduling is concerned but in this instance they may be correct, if for the right reasons or not may yet be debated. Heres why; first you have to ask yourself if those other two machines that are not being used will be in contention at anytime. I.E. will some other department utilize them at the same time. What we are really talking about here is your lowest common denominator, or what is really limiting your capacity. Its your two employees. Vantage does have the ability to reconcile aspects like machine features, they call this capability. this will get you into a whole other world that you may not be ready for. We backed away from this after alot of research. Sounded good on paper, was a nightmare to manage from a scheduling/load standpoint. Start simple, then expand after you have that down pat. You have two employees each working 40 hrs a week in a
department. They can operate one machine at a time 40*(1) + 40*(1) = 80 hrs. Since your operation doesnt sound complex from the way you described it dont make it so. Let your operations people who know the process take care of the execution (which machine to run). Your alternative, list 4 resources each in a different group (unless you want to tackle capabilities) , that must be tied to a calendar of some sort. You still want the system to understand available capacity (as well as you with regards to overloads) so with 4 resources thats (4)*(calendar hrs available on that calendar). This assumes you have a calendar tied to the group, not the resource (which will inherit the group calendar by default). If you still wanted only 80 hours that week your calendar would have to reflect that. Assuming you worked mon - thur in that area you would only have 5 available hours per day. There are alot of variables here, im sure you can extrapolate the intended or
unintended consequnces of each. Sorry for the length, but i love to talk about scheduling.
--- In vantage@yahoogroups .com, "Steve Bedilion" <sbedilion@. ..> wrote:
>
> Sorry for posting this again...but support hasn't been able to shed light on it and I didn't get any feedback the first time i posted, and I would think it's a COMMON scenario...especial ly in todays times.
>
> We have a Bending Department with 4 machines.... but only 2 employees. Each employee can only run 1 machine at a time....so at any given time 2 out of 4 machines can be running. The machines are slightly different, so the Job **must** pick the machine it needs.
>
> If we have a 1 to 1 for workers to machines then its easy...
>
> Resource Group = Bending
> Resource = Machine1, Machine2, Machine3, Machine4
>
> But we want to show "Shop Load" for a Department so you have to take into account that we have 2 employees.
>
> Support's recommendation was as follows...
>
> Resource Group = Bending
> Resource = Person1, Person2
>
> This works perfect when looking at Shop Load...but how to you reference what machine the job needs to run on?
>
> Please help....we are running blind in this department and are having to give out overtime everyday.
>