Scribd
Upload
147 views18 pages

WBS-2-Operations Analytics-W2S1-How-to-Build-and-Optimization-Model

This document describes an optimization model for Zooter Industries to determine the optimal production quantities of two scooter models. It includes: 1) Details on the scooter models, production process, available resources, and the objective of maximizing weekly profit. 2) How to represent the production quantities as decision variables, the profit as the objective function, and resource constraints in the optimization model. 3) How to evaluate whether a given production plan is feasible or optimal by checking if it satisfies the constraints and objective function. The complete optimization model is defined as maximizing the profit subject to constraints on the available production resources. Solver will be used to find the optimal solution.

Uploaded by

rpercor
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
147 views18 pages

WBS-2-Operations Analytics-W2S1-How-to-Build-and-Optimization-Model

This document describes an optimization model for Zooter Industries to determine the optimal production quantities of two scooter models. It includes: 1) Details on the scooter models, production process, available resources, and the objective of maximizing weekly profit. 2) How to represent the production quantities as decision variables, the profit as the objective function, and resource constraints in the optimization model. 3) How to evaluate whether a given production plan is feasible or optimal by checking if it satisfies the constraints and objective function. The complete optimization model is defined as maximizing the profit subject to constraints on the available production resources. Solver will be used to find the optimal solution.

Uploaded by

rpercor
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 18

Week 2: Making Best Decisions in Settings with Low

Uncertainty

 A resource allocation example: Zooter Industries Session 1


 Converting a verbal problem description into an algebraic model:
decisions, objective, constraints

 From an algebraic model to a spreadsheet implementation: optimizing


with Excel Solver Session 2

 Matching demand and supply across space: Keystone Dry Goods


Logistics Session 3

OPS Analytics MOOC, Week 2, Session 1


Zooter Industries: Products, Profits, Demand

 Zooter Industries (ZI) manufactures high-end kick-scooters for the North


American market

 ZI’s main product models are Razor and Navajo, with profit contributions
of $150 and $160 per unit

 At present, ZI’s scooters are so popular that the company can sell all the
units it makes

OPS Analytics MOOC, Week 2, Session 1


Zooter Industries: Manufacturing Process

 The production process for each model includes three main steps:
– frame manufacturing
– wheels and deck assembly
– quality assurance and packaging

 Each unit of the two scooter models requires the following processing
times in these production steps:

Frame Wheels and Deck Quality Assurance


Manufacturing Assembly and Packaging
Model
(hours) (hours) (hours)
Razor 4.0 1.5 1.0

Navajo 5.0 2.0 0.8

OPS Analytics MOOC, Week 2, Session 1


Zooter Industries: Supply Side

 ZI’s capacity available at each production step is shown below for the
coming week

Available Time in the


Production Step Coming Week (hours)
Frame Manufacturing 5610

Wheels and Deck Assembly 2200

Quality Assurance and Packaging 1200

 How many units of each model should ZI produce in the coming week in
order to maximize its weekly profit?
OPS Analytics MOOC, Week 2, Session 1
Assuming Away Uncertainty: Pros and Cons

 The Zooter example treats profit contributions, manufacturing


requirements, supply availabilities as non-random quantities

 If ZI decides to make a certain number of units of each scooter model in


the coming week, it will know for sure
– How much profit it will make
– Whether it will have sufficient supply of each resource

 The “no uncertainty” assumption simplifies the search for the best
production plan

 In practice, it allows us to tackle analytics models with large numbers of


products and resources

OPS Analytics MOOC, Week 2, Session 1


Assuming Away Uncertainty: Pros and Cons

 May be a reasonable assumption when a decision maker has substantial


control over his/her business environment
– Short-term planning
– Longer-term planning when existing contracts ensure stability of prices,
costs, and demand and supply parameters

 May result in problematic recommendations in settings with significant


data uncertainty

 When uncertainty is significant and must be included in the analysis, the


task of finding the best decision may become far more complex

 In Weeks 3 and 4 we will look at how to evaluate choices and make best
decisions in such settings

OPS Analytics MOOC, Week 2, Session 1


Evaluating a Production Plan: Decision Variables

 Before approaching a task of finding the best production plan, or


optimizing production, we must know how to evaluate any given
production plan

 In optimization lingo, the term “decision variables” describes the


quantities that a decision maker can change to achieve a desired
performance.

 In the ZI example, there are two decision variables:


– R, the number of Razor scooters to produce in the coming week
– N, the number of Navajo scooters to produce in the coming week

 A particular choice of values for decision variables is called a “solution”.


For example, R=500 and N=500 is a solution

OPS Analytics MOOC, Week 2, Session 1


Evaluating a Production Plan: Objective Function
 If ZI decides to produce R=500 Razor and N=500 Navajo scooters in the
coming week, how much profit will ZI make in this case?
– Profit (in $) = $150*500*+$160*500 = $75000 + $80000 = $155000

 The “objective” is a performance metric we want to maximize or


minimize. In this example, profit is an objective to be maximized

 For R=500 and N=500, the profit value is $155000. How much profit will
ZI make for an arbitrary pair of values R and N?
– Profit (in $) = 150*R + 160*N

 150*R + 160*N is an “objective function”, i.e., an objective expressed


as a function of decision variables

 $155000 is an “objective function value” (OFV) for solution R=500,


N=500
OPS Analytics MOOC, Week 2, Session 1
Evaluating a Production Plan: Constraints

 If ZI decides to produce R=500 Razor and N=500 Navajo scooters in the


coming week, how much of each resource will it require?

 Required number of frame manufacturing hours:


4*500+5*500 = 4500 – does not exceed 5610 hours available

 In general, for any potential production plan, the required number of


frame manufacturing hours may not exceed the number of hours
available

 In the optimization lingo, we use the term “constraint” to describe this


requirement

OPS Analytics MOOC, Week 2, Session 1


Evaluating a Production Plan: Constraints

 Does the R=500 and N=500 production plan have enough of other
resources to be implemented?

 Required number of wheels and deck assembly hours:


1.5*500+2.0*500 = 1750 – does not exceed 2200 hours available

 Required number of quality assurance and packaging hours:


1.0*500+0.8*500 = 900 – does not exceed 1200 hours available

 A production plan that, like R=500 and N=500, satisfies all constraints is
called feasible

OPS Analytics MOOC, Week 2, Session 1


Evaluating a Production Plan: Constraints

 What if ZI decides to produce R=500 Razor and N=750 Navajo scooters?

 Required number of frame manufacturing hours:


4*500+5*750 = 5750 – exceeds 5610 hours available

 Required number of wheels and deck assembly hours:


1.5*500+2.0*750 = 2250 – exceeds 2200 hours available

 Required number of quality assurance and packaging hours:


1.0*500+0.8*750 = 1100 – does not exceed 1200 hours available

 A production plan that, like R=500 and N=750, violates at least one
constraint is called infeasible

OPS Analytics MOOC, Week 2, Session 1


Evaluating a Production Plan: Constraints

 For a production plan that makes R Razor and N Navajo scooters, how
does one express a constraint on the number of available frame
manufacturing hours?

 In words, we have “number of required frame manufacturing hours may


not exceed the number of available hours”

 Using variables R and N, we can write this statement as


4*R + 5*N ≤ 5610

 In the same way, the constraints on the number of available wheels and
deck assembly hours and the number of available quality assurance and
packaging hours can be written as
1.5*R + 2.0*N ≤ 2200
1.0*R + 0.8*N ≤ 1200
OPS Analytics MOOC, Week 2, Session 1
Other Constraints?
 Numbers R and N must be integer

R, N = integer

 Numbers R and N cannot be negative

R, N ≥ 0

OPS Analytics MOOC, Week 2, Session 1


Searching for the Best Production Plan:
A Complete Model
 Putting the decision variables, objective function and constraints
together, we can express our model as

Maximize 150*R +160*N


subject to
4*R + 5*N ≤ 5610 (frame manufacturing hours)
1.5*R + 2.0*N ≤ 2200 (wheel and deck manufacturing hours)
1.0*R + 0.8*N ≤ 1200 (QA and packaging hours)
R, N = integer
R, N ≥ 0

 We will use Solver to “optimize” this model


OPS Analytics MOOC, Week 2, Session 1


A Comment on Objective and Constraints

 An optimization model can have any number of decision variables and


constraints but it must have one objective to be maximized or minimized

 In practice, there could be a number of quantities, key performance


indicators, that a manager may want to keep track of: profit, cost,
customer service levels, utilization of resources, etc.

 If one of the key performance indicators, such as profit, is selected as


the objective, the rest of the key performance indicators can be used in
constraints

 For example, the model can “maximize profit while making sure that the
resource utilization does not exceed 95%”

OPS Analytics MOOC, Week 2, Session 1


Model Types: “Easier” …

 Zooter model involves only
– constant parameters, like 5610
– products of decision variables and constant parameters, like 150*R and
0.8*N
– adding and/or subtracting the resulting expressions, like 1.5*R + 2.0*N

 Such models are called “linear” and easier to optimize in practice

OPS Analytics MOOC, Week 2, Session 1


… And “Harder”
 Sometimes it is necessary to use models that involve “nonlinear”
expressions of decisions variables, for example, R*N, R2, N/(R+N) or
𝑁

 Nonlinear models are much harder to optimize, especially as the


number of variables and constraints grows

 In the Zooter model, the numbers of scooters produced must be round,


or integer

 In general, imposing such a requirement can significantly complicate


optimization even in linear models, especially in models with large
numbers of variables and constraints

OPS Analytics MOOC, Week 2, Session 1


Additional References

 More on optimization, linear, non-linear models as well as models with


integer variables:
– “Business Analytics” by James R. Evans
– “Spreadsheet Modeling and Decision Analysis: A Practical Introduction to
Business Analytics” by C. Ragsdale

OPS Analytics MOOC, Week 2, Session 1

You might also like