|
|
|
Case Studies
|
|
|
|
CASE #1: Locomotive production planning with AutoSched
|
|
How can we generate and analyze various planning scenarios in 30
minutes, when generating a single scenario required us an
entire week of effort ?
|
Situation:
To build locomotives, from the planning and scheduling point of view, looks very much
like building planes: each order is a unique project even though it is for a
known model. Indeed, all the customers have their own special requirements.
Before AutoSched, generating a 3 month production plan for the main assembly lines and
the production centers was requiring one full week of work for the two planners.
Now, with AutoSched, the same activity is done within 5 minutes! While it was
impossible for them to test different scenarios (i.e. the first plan had to be the
good one) with their original method, the two planners can now concentrate on analyzing
and fine-tuning their plans. In the same way, they have now the flexibility to
regenerate this plan and test scenarios more often and faster, to take into account
for example changes in the orders or production problems. Even though these two
benefits are merely intangibles, the competitive advantage they give to the company is
obvious.
Solution:
How did this all happen? After a week of formal training on
AutoSched, provided
on-site by MultiCIM, and a few weeks of practice, the planners began to create the
model of their facility, focusing first on the main production and assembly line.
Using manual data entry or by importing the information into the spreadsheet interface
of AutoSched, they were able to create the basic simulation model within a couple of
days. This model included the definition of the work centers, a list of models,
their main sub-assemblies and their respective routing and BOM. In close
collaboration with MultiCIM's consultants, this model was then refined to take into
account their unique production constraints and objectives, as well as the Theory Of
Constraints (TOC) approach they were using for scheduling the assembly and supply
lines. Special decision rules were developed to mimic the way this scheduling
had to be done. Finally, custom decision reports (for uploading into the MRP
system, the management, and the shop floor ) were designed in AutoSched to provide the
company with the same type and format of information they were using before
AutoSched. These reports are now automatically generated by
AutoSched.
This took a cumulative total of 10 weeks of work, during which the model was
implemented, tested and validated against different samples of data.
|
|
CASE #2: Coordination of equipment and forge simulation with AutoSched
|
|
How can we achieve a higher productivity and synchronize our forges and furnaces
given our numerous production constraints?
|
Situation:
The rotary forge department of this US company has just doubled its production capacity
by adding a new forge. These two machines are surrounded by auxiliary equipment
(furnaces, saws, cooling baths, conveyors, etc.) and all of these activities have to be
well coordinated to minimize the cost and keep a high level of plant productivity.
A large amount of constraints have to be considered in order to achieve this
objective. Already difficult with a single machine, this task was becoming
impossible to handle without the help of a finite capacity scheduling package.
After unsuccessfully trying a Artificial Intelligence-based package, the company turned
to AutoSched and MultiCIM to implement a viable solution.
Problem:
How can we take into account the wide spectrum of constraints (hammers' limited
capacity, heating/cooling of furnaces, limited quantity of hot boxes for special
material, etc.) and optimize the resources' utilization while keeping the production
costs as low as possible?
Solution:
MultiCIM was hired to develop an AutoSched model of the rotary forge department that
would take into account all the production constraints. This model was to be
integrated with the new Manufacturing Execution System (MES) from which it would
retrieve the information regarding the orders on their way to the rotary forge and the
current WIP. The development focused on the ease-of-use and accuracy of the
model. In order to simplify its daily use and maintenance, the number of routings
was kept to a minimum while the manufacturing information was passed together with the
order information received form the MES. The unique loading scheme of the furnaces
has been reproduced in the model and customized decision rules were developed for each
piece of equipment. These rules were built by MultiCIM's consultants after
conducting a series of interviews with the shop engineering and scheduling team.
The flexibility of AutoSched allowed us to fine-tune and improve these rules while
testing the model under various production scenarios. These decision rules are
capable of dynamically balancing account local constraints (e.g. current hammer set,
current capacity available, etc.) together with global information (e.g. quantity of
pieces still remaining to be forged on the current hammer set, contents and temperature
requirements of the material in support furnaces, etc.).
According to the company, the model, which is now used daily since 1995, allowed them
to generate better schedules faster than the expert human scheduler would have been
able to do.
|
|