Tunnel Vision

Last month’s column

described the structure of MRPII systems. The column

ended with the statement that such systems suffered from

being too narrow and manufacturing oriented. In this

column, I intend to justify the statement and lead the

way for deriving the structure of an ideal ERP system.

A major problem with pure

MRPII systems was that they were not integrated with the

financial systems used in the organization. Implementers

were prone to treating these systems as separate

subsystems within the same organization. This created a

problem at the point of intersection of the manufacturing

and financial domains. Top management was given a

production and manufacturing plan and asked to approve

it. To approve the plan, top management needed

information from the financial system regarding

receivables and the like. In addition, they wanted

information from the sales systems on current sales level

to cross check with the production plan. But, the

financial and sales systems were treated as separate

systems run by separate teams, creating artificial

barriers to the smooth flow of information.

The two systems often had

widely different interfaces, and generated information in

formats not compatible with each other. Consequently,

information had to be reprocessed and could not really be

considered online. This leads us to one of our first

requirements for a good ERP system–all subsystems

should act as components of a more unified system.

The next problem with MRPII systems

was that they were essentially single manufacturing

centre systems. They assumed that all in-house production

took place under one roof. The models used in these

systems were simply not extensible enough.



For instance, a large company could have one large
central plant producing core components, and multiple

small assembly plants spread geographically. It was quite

a pain extending the MRPII system to cover this

situation. The classic MRPII system would require that

all information be aggregated at the central plan,

production planning and scheduling done, and the plans be

distributed back to the smaller assembly plants.

Classical MRPII systems

also did not pay enough attention to the management of

the total supply chain. MRPII was manufacturing oriented

to an unhealthy degree. To drive home this point, let us

take a look at some hard examples.Given a demand projection,

managements are often faced with a multitude of options.

For instance, the company might be facing a stock

shortage in one state while having excess stock in a

warehouse in another state. Consequently, one is faced

with a manufacture versus transship decision which

requires inputs from the manufacturing, finished stock

inventory, and logistics modules.

Another example relates to

production requirement’s forecasting. MRPII normally

relies on simple statistical models to forecast demand.

Current day demand forecasting makes detailed use of

current sales data, and tries to derive future demand

from this data and current market conditions. This

process encourages the use of human inputs. One possible

way of doing this is to help field sales-staff give their

assessment on how much each key account will purchase in

the next quarter, and then aggregate these estimates.

All this helps us refine

our expectations from enterprise-wide software. We now

expect that ERP software should enlarge its focus from

manufacturing systems alone, and help us take a more

holistic look at the organization. And that’s what I

want to discuss next month in the concluding column of

this series.

The Bottomline—the

transition from MRPII to ERP involves moving from a

manufacturing-only mindset to a total-value-chain

mindset.