REPLY CFP INFORMS: Behavior dynamics in operations (SD6425)

[SDMAIL Jean-Jacques Laublé]

Posted by  Jean-Jacques Laublé <jean-jacques.lauble at wanadoo.fr>

Hi Kazem

I am not doing research about operations, but adjusting modelling work with
ground floor necessities has been one of my principal difficulties.
All of my modelling have been oriented towards ground operations so far.
There are two ways of adjusting the necessity of a high level of aggregation
with ground operation. One can use both at the same time.

Taking a concrete example suppose that the subject is a factory making 100
products and wanting to plan how much of each products will be produced the
next semester.
A high level model exists but where production is not disaggregated at all.
The model does not precise the exact production of each product.
Disaggregating may be very time consuming eventually impossible, due to
the lack of disaggregated data for example.
At the operational level people need to know what production is planned for
each product.

Suppose that the model advice is to increase production by 10%.
At the operational level, people will find this impossible to realize
concretely and uniformly for each product.

'The production of XYZ can be increased but it is absurd to increase
production of ZZZ.'
by example.

One solution is to disaggregate a little the model for instance down to 10
families of product and ask the operating agents to aggregate the 100
products into 10 families having the best common characteristics.

For example one of the characteristic can be the possibility to increase
concretely or not the production.
The choice of the families will be critical.

Once this is done, there will always be some dissatisfaction at the ground
level, but it will be necessary to explain that the objective is to find a
global amelioration even if in some cases the detailed adjustment is not
correctly done and that it can always be corrected later on.

The other solution that can be mixed with the first one and if, inside the
family, there are still insupportable policies, to define factors that are
accepted and that can modulate the detailed policy, with positive or
negative adjustment, trying to maintain the overall family policy to the
level advised. The family can be the whole products if disaggregating the
model is too complex. There must be some arbitration to avoid exaggeration.
For instance at the ground level, people overestimate the effect of detailed
policy upon the overall results.

One of the great advantage of operations over strategy is that one can
adjust with the time, policies and that one must understand that adjustment
takes time and learning from experience, but that due to the relative short
cycle of decisions, any bad decision can be relatively quickly corrected.

Combining both approaches is what I have done so far and it works well in my
case.
Regards.
Jean-Jacques Laublé 
Posted by  Jean-Jacques Laublé <jean-jacques.lauble at wanadoo.fr>
posting date  Fri, 20 Apr 2007 17:18:21 +0200