PPDS optimization

[Former Member]

friends,

we have a complex scenario and need help solving this :

scooter manufacturing is planned and planned orders are created by production planning run ( can be a direct forecast at the plant or distribution forecast from other plants). Now I want to schedule these orders in such a way that I want to group them by colour and the plant to which it is going. Secondly I want to make sure that I optimally utilize my space in container which goes to the plant periodically.

Can anyone explain the steps in getting this activity done?

Thanks

Prod_Planner

[Former Member]

Thanks for the reply Virender.

I like your idea of breaking it down into 2 pieces. Let me explain it lil more in detail :

for objective 'A', as suggested by you, we can use setup matrix to come up with the sequence. Now for 'B' I have following requirements

- have to max utilize my container

- In doing so we can pull in more orders from future dates to fill the container

Example : AFter optimization (A) lets say we have following sequence of scooters to be produced:

XXXYYYZZ. X occupies 0.5 container, Y occupies 0.3 container and Z occupies 0.2 container. Now plant1 and plant2 needs XYZ (i.e. 0.50.30.2 = 1 full container) and plant3 needs XY (i.e. 0.5 + 0.3 = 0.8 of container) expecting that I have enough capacity, now I need some way to pull in that extra scooter for plant3 so that I fill my container going to plant3.

I may not want to run SNP optimizer at this point of time in the planning process. I am sure that 'B' cannot be solved as part of PP/DS. Can we by any chance have some custom process in place to achieve 'B'. I think we may try to use DSP board to do manual workaround, but I may need an automated way to do this.

Pls pour in your ideas.......

[Former Member]

HI Planner

if i get your requirement right you mean to say u have one Production plant and three DC.

@ the Production plant you have got a Production sequence of XXXYYYZZ

PLANT1(DC1) Needs XYZ = 1 container

PLANT2 (DC2) Needs XYZ = 1 container

PLANT3 (DC3) Needs XY =0.8 container

But you want to supply 1 continer to DC3 which is 1 full container and you want to pull data for filliong that truck.

Fist point is a demand will not be pulled if you dont have receipts in place.

One way of looking at it would be have production plans in place and then try to pull the demand from future to

optimize the utilisation of container.

in Ur case if the capacity is not a problem the demand has to be inline with the container size.

I would suggest you put a transportation resource in the tlane which has a capacity of 1 container.

Then with a higher SS days at plant 3 take a heuristic run and then capacity levellling at the Transportation resource

level. This will make the STR from plant 3 on the production plant equal to 1 container every day.

Take a PPDS heuristic on the demand to generate planned orders and then the Sequencing heuristic.The solution can also be

modified to the number of containers available everyday. You can put the capacity accordingly in the capacity profile of the resource.

The STR in the initial horizon will be perfect as the SS days maintined will lead to higer STR in the initial buckets which will

be leveled across certain horizon.

-Try this and let me know if it helps u in some way

Viren

[Former Member]

HI Planner

it will be better if u break ur problem into two parts ,

A Grouping of same color Scooters together.

B Optimal utilisation of container

For A you can try a heuristic as well as optimisation technique using set up groups.

Each color can have a unique set-up group defined in the PPM.

An optimizer as you would already know would try to minimise set up times and give

a solution for minimum changeover.

If you dont have a PPDS optimizer installed u can also do it with heuristic (schedule sequence manually).\

The interactive screen which opens up you can sort the orders according to set up gp. (Colors).

ByOptimal utilisation i think u mean to say whatever is the day's production should be

equal to the container size to be shipped.

First thing is the demand has to be inline to what ur requirement is, atleast it should be more than the container size

for the time period of planning. in a way you would want to pull the demand from future to optimize

your container size. There are various options to do it

1. SNP optimizer - put a secondary bucket resource in PPM with min. utilisation and capacity equal to the container and choose the

capacity variant 2. This will assign a notional cost for under utilisation of resource and will try to fetch the demand from

future to complete the container size.

2. Try brining the demand from the T lane resource in such a way that the net demand falling in a day is equal to the container size and can be produced also.

3. Try deployment and TLB in which u can pull the demand from future and make STO equal to the truck size but it might lead to some residual receipts unconverted.

- Virender