Skip to main content
There are more than one occasions that I have come across that our users define one physical machine in 2 sites expecting to plan shop order operations seamlessly. But this set up sometime leads to very confusing consequences.



Example :



Company : A

Site : X

Site : Y



Expensive CNC machine is installed in Site X ( defined as a manufacturing work center in Site X as "CNC-X").

But this usually runs under capacity ( lets assume that both sites runs on 8 hr calendar). So The company decides that they can put some shop orders registers in Site Y to be done in the under utilized work center in Site X.



Therefore it is decided to define the work center in site Y as " CNC-Y" in IFS.



Now lets assume that the manufacturing planner has access to both site X and site Y shop order operations.



If you run CRP, since in IFS , CNC is 2 distinct work centers, the capacity available is 8 hr each. Which is not really the case. For both sites , the available hours are 8 in total per day.

In such setup planners need be vigilant to not to overload the "common " machine for both sites. Manufacturing visualizer and loading Advance Planning Board with resource break reflecting loads on the "other" site might help. But in any case, 2 work centers represent one single physical CNC machine which is not true. Generic resoruce may be a ( distant ) complete solution.



Q1. Is it a common practice that the assets are defined in that way?

Q2: In such setup, what are the approaches that planners has take to take best use of IFS Planning functions?

Q3. Anyone has any suggestions or best practices to recommend for such situations ?



your comments are appreciated.
Interesting scenario. Let's wait to get feedback on whether this is a common scenario!



Solution wise it could probably be handled similar to as it is handled for labor. I.e. one person (resource) could be split among several labor classes (even on different sites) using a % factor. So one machine resource could be split among work centers using the same logic.
Thanks @Björn Hultgren ,

We have done this in MPB ( Maintenance Planning Board ) where work orders from different sites users same maintenance organizations and crafts. As you said, the question is how common is this practice.
Thanks @Björn Hultgren ,

We have done this in MPB ( Maintenance Planning Board ) where work orders from different sites users same maintenance organizations and crafts. As you said, the question is how common is this practice.




That may work for maintenance planning but what about normal manufacturing planning. As I have learned IFS lo these past 20 years, a site is a separate operational entity and its workcenters are limited to work to be performed within that site. I would be interested in seeing how this may be accomplished with a resource (machine) that is common across two sites.
@Vernon Anderson

Hi Vernon, Thanks. That is what we are also would like to see (

How common users wants to use a same physical machine to perform orders coming from 2 different sites). In the specific case I mention, the sites are defined to seperate "work" but they want to use the excess capacity of the machine in one site to perform "work " from the other site.
Perhaps the easiest way would be to give them access to the manufacturing visualizer in both sites as a custom RMB from one work center to the other?
@Vernon Anderson

Thanks, That is one approach of course. But Lets say the user is looking at CNC-X and CNC-Y load graphs, say machine op load in std hrs. For CNC-X he will see a capacity line of 8 hrs. which is not true. At the same time , 16 hrs availability line in CNC-X, and CNC-Y is also not true. Because the physicle CNC has only 8 hrs total. I actually tried to design a good "work pattern" for the planner to plan this common CNC machine using visualizer (and using legend turn on off, publisher subscriber, and group by functionalities )but could not come out something robust and easy to use. This is purely a problem of how business is mapped. IFS expect machine and loads to be in same site in an manufacturing environment. Let me know if you come out of some work pattern which is seamless. - Thanks

I’ve encountered this scenario multiple times .Even across multiple companies.

Example: weaving,Knitting mills ,Tea Packaging . -Common factor -Single planning department

the first two can have 100s of machines at a single site.

Like Saman has mentioned ,The same machine is registered in all sites as work center resources.

As far as I know none of them uses CRP. what they do is look at the machine load grouped at machine level .

as to identify the machines real location a custom filed is used to identify the actual ‘Site’


Interesting scenario. Let's wait to get feedback on whether this is a common scenario!

Solution wise it could probably be handled similar to as it is handled for labor. I.e. one person (resource) could be split among several labor classes (even on different sites) using a % factor. So one machine resource could be split among work centers using the same logic.

Hi Bjorn,

If the planners function is to balance the production load for any given time frame ,I don’t think a fixed capacity share between a sites is a good solution.

eg:

This week CNC-X  is utilized solely for production of the site X &  next week it may be utilized for shop orders in both sites.

 


Interesting scenario. Let's wait to get feedback on whether this is a common scenario!

Solution wise it could probably be handled similar to as it is handled for labor. I.e. one person (resource) could be split among several labor classes (even on different sites) using a % factor. So one machine resource could be split among work centers using the same logic.

Hi Bjorn,

If the planners function is to balance the production load for any given time frame ,I don’t think a fixed capacity share between a sites is a good solution.

eg:

This week CNC-X  is utilized solely for production of the site X &  next week it may be utilized for shop orders in both sites.

 

True @Bhathiya Ranasinghe  My point was rather that when evaluating that machine resource (work center resource) load vs capacity of course load from any site would be presented against the resource full capacity. But if that resource would be a member in several machine groups (work centers), evaluating the load vs capacity on one of those machine groups would include only the part of the capacity from the shared machine resource as specified in the % in group.


Hi mate, I have seen this happening first hand at least in two instance when I was implementing IFS (back in the day); but more significantly, with one customer who was one of the biggest value added tea manufacturers.

 

From the back of mind, here are some facts I gathered:

 

  1. What do they do: Pretty much what you explained. Using a common set of physical machinery across two companies. The assets are registered under the main company, and the work centers too. But during the low tide, it will be running manufacturing from company B, along with orders from company A.

 

  1. What’s their IFS manufacturing environment: App 7.5, DOP, MRP, APB, No CRP, No Visualizer (wasn’t heard of it back then)

 

  1. Challenge with doing this in IFS: My initial check was to see the possibility of using inter-site/ Distribution Order  functionality, so the company B could send a proper inter-site CO that initiates an SO which would get allocated along with company A’s orders on one resource, and then scheduled with APB. This seemed like proper way with IFS standard solutions, but it was not gonna work practically due to the complexity of supply chain, the amount of human intervention required, and just too many clicks for what they deemed a “simple problem”.

 

Here are two things I tested with as potential solution:

 

  1. Solution 1 (the more ambitious one):
    1. Set up a dummy site X in either company A or a dummy company C.
    2. This site lists all manufacturing resources and acts as the “combined view”. All inventory, financial transactions etc are blocked and only allows creating manufacturing constraints including shop orders. APB runs only on this dummy site. The other (main) sites runs infinite scheduling.
    3. How it works is, when company A or B creates shop orders on their respective sites, a custom batch process will create a replica of those orders in site X. This process will run as a background job to free up the client. Now while the original SO is doing all financial and inventory transactions, its replica order in the dummy site purely does the finite scheduling. Hope it makes sense.
    4. Custom events are in place to propagate any applicable changes from the original SO to its replica (ex. Order qty changed, SO gets cancelled, gets parked, split, added operations etc.).
    5. While this approach involves a lot of custom configurations (means more prone to bugs, taking up database CPU etc), it will work with complex shared resource scenarios like Saman mentioned, as two replica orders are on the same dummy site and fighting for the same resource – which will be taken care of by APB.

 

  1. Solution 2
    1. Using resource breaks. Following your example, CNC-X and CNC-Y are added under the same department or other means to let IFS know they are physically the same resource. When, say, CNC-X gets loaded 3 hours from 8AM to 11AM, a custom event creates a resource break in site Y to block out CNC-Y between 8AM and 11AM. A description on the resource break will tell the planner this was actually triggered by site X. APB will schedule the resource break in. This leaves site Y only 5 hours remaining to book on the shared resource.
    2. Custom events should be in place to replicate changes back and forth on the resource breaks.

 

So which solution worked? Not sure. Think I left before they concluded their back and forth arguments, or they had bigger issues to fix. They might be still doing it manually which often is what works in the end.

 


For us...because we have a need to maintain inventory accurately in both sites, not to mention cost in both sites, material drive, etc.  The best way we have found is to use the out of the box Intersite transactions, but then automate as much as possible using Order Point settings, auto released purchase orders, auto released customer orders in the receiving site, auto released shop orders based on the MRP demands.  IFS sees tons of transactions, but the humans do relatively little key punching in comparison.  We don’t use CRP and we let the site with the machining center manage the workload based on internally generated demand as well as the direct external demand for the cross-site.


Reply