INTRODUCTION TO

FORECASTING, PLANNING , MRP,

MRP II AND ERP
 Said to be art and science of
predicating the future

Can be based on historical data and

mathematically extrapolated for
prediction (Modelling)

Possibly subjecting or intuitive

FORECASTING prediction.

Could be based on customer intent.

Or a hybrid of the above.

 Term

FORECASTING
HORIZONS As a guide:
• Short Term – up to a year normally less than 3
months.

• Medium Term – 3 months to 3 years

• Long Term – Greater than 3 years

 TIMING DIFFERENCE

• Medium to Long term typical consider plant / product / process development

• Short term typically utilises moving averages, exponential smoothing and
trends.
• Short term forecasting is often more accurate, as less assumption required and
time line shortened to prediction, reducing longer averaging and variation
impact.
 Economic

FORECAST Technological
TYPES

Demand
 Plan for business expenditure and resultant return on
1 investment.

Understand what investment maybe required to achieve

2 goalsof business.

WHY 3 Allow supply chain to prepare for demand and their

subsequent forecasts.
FORECAST?
4 Allow for resource to support investment to be put in place.

Understand the potential future P/L, EBITDA, Balance Sheet

5 position of the business.

6 Understand capacity requirements.

 1 Define what the forecast is for

2 Identify items to be forecast

3 Define time scales, short , medium, long

STEPS TO
FORECASTING 4 Select models to be used, e.g. moving average, regression.

5 Collect data.

6 Build the forecast

7 Confirm and produce the forecast information

 Two general approaches:

Qualitative (quality of something,

APPROACHES opinions, judgements)
• Jury of executive opinion – high level exe.
• Delphi method – decision makers, staff,
respondents,
• Sales force composite – salesperson
estimates forecast
• Market survey
 QUANTITATIVE APPROACHES

• Quantitative (measuring of something, actual data,

quantifiable)
o Time Series Forecasting Models, (Assumption future based
on past, trend, seasonality, cycles, random variations)
 Naïve approach, based on previous period, eg last
months sales will be this months.
 Moving Averages, good if demand is relativity stable
Moving Average = Sum Demand n Months / n
(number of periods). Lagging occurs if trend evident.
 If trends/pattern evident weighting can be used
for emphasis.
 Weighted moving average = Sum ((Weight for
period n)(Demand in period n)) / Sum Weights
 Increasing size of n reduces sensitivity to data changes.
 Not adept at modelling trends/patterns.
 Needs past data for calculation.

(Heizer, J et al. 2017)

 QUANTITATIVE APPROACHES

• Quantitative (measuring of something, actual data, quantifiable) • To adjust for trends there is one approach: Second order or double
smoothing,
 Exponential smoothing (first order), weighted moving  Forecast including trend (FITt) = Exponentially smoothed
average method. Does not require extensive past data. forecast average (Ft) + Exponentially smoothed trend (Tt)
 Ft = Ft-1 + a(At-1 – Ft-1) • Ft = a(Actual demand last period) + (1 - a)(Forecast last period +
Trend estimate last period) or:
 where F t = new forecast
Ft = a(At - 1) + (1 - a)(Ft - 1 + Tt - 1)
 Ft-1 = previous period’s forecast
Tt = b(Forecast this period - Forecast last period) + (1 - b)(Trend
estimate last period) or:
 a = smoothing (or weighting) constant (0 ≤ a ≥1)
Tt = b(Ft - Ft - 1) + (1 - b)Tt - 1
 At-1 = previous period’s actual demand
where F t = exponentially smoothed forecast average of the data
 Appropriate values of a need to be used for accurate series in period t
forecasting.
• T t = exponentially smoothed trend in period t
 If average is likely to change then use higher values of a
• A t = actual demand in period t
 If not then low values of a. • a = smoothing constant for the average (0 ≤ a ≥1)
• To adjust for trends there is one approach: • b = smoothing constant for the trend (0 ≤ a ≥1)
 Forecast including trend (FITt) = Exponentially smoothed
forecast average (Ft) + Exponentially smoothed trend (Tt)
 SEASONAL
FORECASTS
FORECAST ERROR
 TREND PROJECTION

• Trend projection based on a linear model.

• Fits a line of best fit to a series of historical data points
then extrapolates for medium / future projections.
Applying the least-squares method:
• A least-squares line is described in terms of its y -
intercept (the height at which it intercepts the y -axis)
and its expected change (slope). If we can compute the
y -intercept and slope, we can express the line with the
following equation:
• ŷ = a + bx
• where ŷ (called “ y hat”) = computed value of the
variable to be predicted (called the dependent variable )
where b = slope of the regression line
• a = y -axis intercept ∑ = summation sign
x = known values of the independent variable
• b = slope of the regression line (or the rate of change y = known values of the dependent variable
x = average of the x -values
in y for given changes in x ) y = average of the y -values
n = number of data points or observations
• x = the independent variable (which in this case is time) We can compute the y -intercept a as follows:
a = y - bx
 TREND EXAMPLE

Assumptions:
1. We assume a linear trend, if not then other trend analysis needed.
2. We can only predict a small number of periods ahead, otherwise statically inaccurate.
3. Data points around the least squares are assumed to be random and with a normal distribution,
 LINEAR REGRESSION

• Associative Models
(accounts for variables
that might affect
quantity)
• Linear regression
 Substitution of the axis ŷ
with a dependent variable
( the factor that you wish
measure agianest) and x
axis with a independent
variable (the other
variables that could be
measured)

(Heizer, J et al. 2017)

 STANDARD ERROR OF ESTIMATE

Standard deviation of the regression, using SD calculation. Therefore SD $306K in Sales

 CORRELATION

• One method to try and understand a relationship is Correlation, the

coefficient of correlation

Note: Correlation does not always imply causality

(Heizer, J et al. 2017)

 MRP, MRP II, ERP

Manufacturing Requirements Planning

Manufacturing Resource Planning (MRP II) Enterprise Resource Planning (ERP)
(MRP)
• 1970’s onward with greater computing • Grew out of MRP in 1980 with improved • Similar to MRP II but on a company wide
power, integral is the Bill of Materials computing a communications, allowed scale with differing departments and
(BOM) and Master Production Schedule what if scenarios and exploration of further enhanced with improved web
(MPS) – demand information. consequences of change. communications. Collaborative commerce
• MRP II coined by Oliver Wight, one of the or c-commerce will allow full exploitation.
founders of MRP. Defined MRP II as ‘a
game plan for planning and monitoring all the
resources of a manufacturing company:
manufacturing, marketing, finance and
engineering. Technically it involves using the
closed-loop MRP system to generate the
financial figures.’
 ERP
RELATIONSHIPS

Slack, N. et al. 2007

MRP REQUIREMENTS

(Heizer, J et al. 2017)

 MASTER
PRODUCTION
SCHEDULES

• Consist of level master

production schedule, levelling
of demand, smoothing.
• An available to promise
schedule, eg still available for
production.
• A Master schedule also
accounts for a number of
products with common
components and can
incorporate a number of
products

(Heizer, J et al. 2017)

 BOMS

• BOM (Bills of Materials) or organised commonly into hierarchical structures.

• There lowest level stops when the component is not made by the company.
• Dependent upon the parts required and the design of the product, BOM
structure can vary. E.G.

Slack, N. et al. 2007

 ‘T’ Type – Raw
materials small,
‘A’ Type – Limited
wide range and
product range
customisable end
products.

COMPONENT
STRUCTURE ‘X’ Type – Small
number of
‘V’ Type – Typically module designs
petrochemical then wide end
industry product range
typically
Automotive.
 MRP II

• Note Requirements change to Resource in

MRP II.
• With Resource we can track and manage
for example:
Money, Energy, Scrap, Packaging usage,
Environmental impacts.
• Nearly all MRP Systems are closed loop,
that is they feedback into other systems,
Purchasing, Capacity Plans, Warehouse
Management, etc.

(Heizer, J et al. 2017)

 Overlapping, which reduces the lead time, sends pieces to
the second operation before the entire lot is completed on
the first operation.

Operations splitting sends the lot to two different machines

for the same operation. This involves an additional setup, but
SMOOTHING results in shorter throughput times because only part of the
TACTICS lot is processed on each machine.

Order splitting, or lot splitting , involves breaking up the

order and running part of it earlier (or later) in the
schedule.

Slack, N. et al. 2007

 ERP

• Whole System Integration

• I 4.0 will highly likely extend this
integration.
• Probable requirement for RFID,Vison
system, AR,VR and AI with associated
mobile applications.

(Heizer, J et al. 2017)

 MRP systems require
accurate BOM’s if errors Change control needs to Requires constant
occur then these can be be tight controlled, management and
multiplied rapidly through avoiding multiple copies. oversight.
the hierarchy.

The support structures

for the ERP system can
ERP Systems can be Often said to give
cost more the ERP
expensive in time and negative return on
DISADVANTAGES cost to deploy. investment.
software (Consultants,
hardware, networks and
support systems)

Can sometimes require

organisations to change
their mode of operation
to fit the ERP system.
 Considered
Their use and
Forecasting
applicability
methods

IN SUMMARY
Their
differences,
MRP, MRP II & application,
ERP requirements
and
disadvantages.
 REFERENCES

Slack, N., Chambers, S. and Johnston, R., 2007. Operations management. Pearson
education.
Heizer, J., Render, B., Munson, C. and Sachan, A., 2017. Operations management:
sustainability and supply chain management.