You are on page 1of 19

This article was downloaded by: [129.93.16.

3] On: 23 October 2015, At: 04:01


Publisher: Institute for Operations Research and the Management Sciences (INFORMS)
INFORMS is located in Maryland, USA

Interfaces
Publication details, including instructions for authors and subscription information:
http://pubsonline.informs.org

Statistical and Optimization Techniques for Laundry


Portfolio Optimization at Procter & Gamble
Nats Esquejo, Kevin Miller, Kevin Norwood, Ivan Oliveira, Rob Pratt, Ming Zhao

To cite this article:


Nats Esquejo, Kevin Miller, Kevin Norwood, Ivan Oliveira, Rob Pratt, Ming Zhao (2015) Statistical and Optimization Techniques
for Laundry Portfolio Optimization at Procter & Gamble. Interfaces 45(5):444-461. http://dx.doi.org/10.1287/inte.2015.0802

Full terms and conditions of use: http://pubsonline.informs.org/page/terms-and-conditions

This article may be used only for the purposes of research, teaching, and/or private study. Commercial use
or systematic downloading (by robots or other automatic processes) is prohibited without explicit Publisher
approval, unless otherwise noted. For more information, contact permissions@informs.org.

The Publisher does not warrant or guarantee the article’s accuracy, completeness, merchantability, fitness
for a particular purpose, or non-infringement. Descriptions of, or references to, products or publications, or
inclusion of an advertisement in this article, neither constitutes nor implies a guarantee, endorsement, or
support of claims made of that product, publication, or service.

Copyright © 2015, INFORMS

Please scroll down for article—it is on subsequent pages

INFORMS is the largest professional society in the world for professionals in the fields of operations research, management
science, and analytics.
For more information on INFORMS, its publications, membership, or meetings visit http://www.informs.org
Vol. 45, No. 5, September–October 2015, pp. 444–461
ISSN 0092-2102 (print)  ISSN 1526-551X (online) http://dx.doi.org/10.1287/inte.2015.0802
© 2015 INFORMS

Statistical and Optimization Techniques for


Laundry Portfolio Optimization at
Procter & Gamble
Nats Esquejo
Procter & Gamble, Newcastle-Upon-Tyne NE27 0QW, United Kingdom,
esquejo.nl@pg.com

Kevin Miller, Kevin Norwood


Procter & Gamble, Cincinnati, Ohio 45202
{miller.kp@gp.com, norwood.kt@pg.com}

Ivan Oliveira, Rob Pratt


SAS, Cary, North Carolina 27513
{ivan.oliveira@sas.com, rob.pratt@sas.com}

Ming Zhao
Interfaces 2015.45:444-461.

Department of Decision and Information Sciences, Bauer College of Business, University of Houston,
Houston, Texas 77204, mzhao@bauer.uh.edu

The Procter & Gamble (P&G) fabric-care business is a multibillion dollar organization that oversees a global
portfolio of products, including household brands such as Tide, Dash, and Gain. Production is impacted by
a steady stream of reformulation modifications, imposed by new-product innovation and constantly changing
material supply conditions. In this paper, we describe the creation and application of a novel analytical frame-
work that has helped P&G determine the ingredient levels and product and process architectures that enable the
company to create some of the world’s best laundry products. Modeling cleaning performance and other key
properties such as density required P&G to develop innovative quantitative techniques based on visual statisti-
cal tools. It used advanced mathematical programming methods to address challenges that the manufacturing
process imposed, product performance requirements, and physical constraints, which collectively result in a
hard mixed-integer nonlinear (nonconvex) optimization problem. We describe how P&G applied our framework
in its North American market to identify a strategy that improves the performance of its laundry products,
provides targeted consumer benefits, and enables cost savings in the order of millions of dollars.
Keywords: pooling; blending; optimization; response surface; design of experiments.
History: This paper was refereed.

P rocter & Gamble (P&G) laundry products are


global household brands that include Tide, Dash,
and Gain, and are offered in several physical product
Traditional formulation approaches involve simpli-
fying the problem, hypothesizing a solution, physi-
cally creating and testing prototypes, analyzing results,
forms, including powders, liquids, pods, tablets, and and iterating the results until various objectives are
bars. These products are manufactured in more than met. Physical prototyping can be expensive and time
30 sites and sold in more than 150 countries world- consuming, resulting in slow and costly iteration
wide. The design of laundry-product formulations cycles; as a result, these traditional approaches no
(i.e., ingredient composition of chemical mixtures) has longer meet today’s needs.
become more complex over the years because of chal- P&G’s research and development organization is
lenges, such as product-portfolio expansion, rapidly at the forefront of the development and adoption of
changing ingredient costs and availability, and increas- modeling tools that enable the company to make better
ing competitive activity. The pace of change is fast and decisions on product formulation, processing, and man-
increasing. ufacturing. These include empirical, first-principles,
444
Esquejo et al.: Laundry Portfolio Optimization at P&G
Interfaces 45(5), pp. 444–461, © 2015 INFORMS 445

and semi-empirical models that predict chemical reac- products as a group of products with a common set
tions during manufacturing, in-use physical prop- of characteristics. In this paper, we define portfolio as
erties of the product, technical performance of the set of formula-unique powder laundry detergents
the product, and even consumer acceptance rates. manufactured in our North American site.
These tools enable researchers to instantly predict a Figure 1 illustrates the mixing architecture and
product’s physical properties and performance, inte- problem structure of the laundry detergent blending
grate models, and balance production trade-offs using process. A large portfolio of products is created from
a variety of predictive and prescriptive capabilities. a relatively small number of intermediate batches (i.e.,
Until recently, the complexity of laundry-formulation 1 to 8 in Figure 1); an intermediate batch, also called
and manufacturing processes limited us to consider an intermediate, is a mixture that is shared by var-
reformulating only a single product at a time; how- ious finished products. Each product is created by
ever, breakthroughs in mathematical optimization blending a portion of its mixture from exactly one
technology have made possible system-wide portfolio intermediate batch with as many finishing additives
reformulation. This is critically important because it as required. Intermediates and finished products are
permits us to model and optimize product differentia- chemical mixtures of one or more ingredients or fin-
tion within a portfolio and consider sharing common ishing additives. Ingredients or finishing additives are
materials within the manufacturing process. In this sourced in the form of chemical mixtures, which we
paper, we present the scope of laundry-portfolio mod- refer to as premixes. The ingredient composition of
Interfaces 2015.45:444-461.

eling and optimization at P&G, the creation of capa- each premix is given, whereas the proportion of pre-
bilities we developed to address this scope, and its mixes to be combined to produce a desirable mixture
application to innovate P&G’s North American pow- of ingredients must be specified (as a decision vari-
der laundry portfolio. able). Costs are specified at the premix level, whereas
product properties are determined by the ingredient
composition.
Problem Definition and Challenges The goal of production is to minimize portfo-
The P&G North American laundry detergent busi- lio annual material spend across a network, which
ness comprises three product forms: powders, liquids, currently includes about 40 products and up to
and pods. Powder detergents, which generate annual 40 ingredients; material costs typically account for
sales of several hundred million U.S. dollars, are a approximately 60 percent of the total cost of produc-
critical part of P&G’s North American business. Even tion. P&G imposes many constraints to ensure that its
as we focus on powders as the primary application, targeted levels of quality and manufacturing feasibil-
the framework for these tools can (and must) be eas- ity are achieved. These include requirements for stain
ily extendable to other forms. Therefore, although we removal and whiteness performance, material bal-
focus on the powder problem in this paper, we note ance and density of intermediate batches and product
that the liquid form is a simplified version of this mixtures, manufacturing site (i.e., plant) throughput,
problem. water content, and raw material usage. Decisions to
Laundry-product formulation can occur in one or be made include: assignments of products to interme-
more manufacturing sites to supply multiple markets. diates, intermediate-proportion contributions to each
Identical product formulations are commonly made product, mixture compositions of intermediates, and
in three or four different sites to fulfill the demand additive proportions in final products. In addition,
of an entire region, such as Western Europe or North for laundry detergent powders, intermediate batches
America. Because each manufacturing site defines its must conform to unique evaporation rules that make
own set of products, the possibility exists that 80 per- the problem more complex. Making the intermediate
cent of the products produced in two different man- batch requires mixing ingredient premixes in a slurry,
ufacturing sites may coincide, whereas the remaining and then evaporating the excess water to form a free-
20 percent are small-volume formulations that only flowing powder, which is mixed with finishing addi-
one site supplies. We typically refer to a portfolio of tives to create the final product.
Esquejo et al.: Laundry Portfolio Optimization at P&G
446 Interfaces 45(5), pp. 444–461, © 2015 INFORMS
Interfaces 2015.45:444-461.

Figure 1: (Color online) The production of laundry detergent mixtures creates a blending process.

To formalize a solution to this complex problem, we polynomial functions that capture the two perfor-
separate the analysis into two categories: predictive mance qualities of a mixture: stain removal and
models and optimization. Predictive models are used whiteness.
to quantify the various relationships within the system, Empirical models for stain removal and whiteness
and optimization incorporates these predictive mod- were created using experimental design procedures,
els into the mathematical formulation to determine the an efficient means of model creation for controlled
ideal values of decision variables. Next, we provide experiments (Box et al. 2005, Kutner et al. 2004), using
details about each component of the problem. JMP software for the design and analysis. Figure 2
shows an example of an experimental design for a
Predictive Models three-variable model for a coffee stain. The table in
Predictive models are either empirical or semi- the figure lists the set of 16 test treatments we ran
empirical in nature. Empirical models are third-order and the associated coffee-stain response. The image to

Variable 1 Variable 2 Variable 3 Coffee Coffee SRI


SRI 95.0
1 0 100 100 88.56 Variable 1 Variable 2
40 60 80 100 100 80 60 40
92.5
2 50 50 0 81.96 0 20 20 0
90.0
3 100 50 50 85.84 100 100
87.5
4 0 100 0 79.65 82.5
5 0 0 0 86.20 80 80
80.0
6 100 100 100 94.02
Variable 3
Variable 3

60 60
7 100 100 0 83.46
8 50 100 50 85.21
40 40
9 100 0 0 83.48
10 100 0 100 91.88 20 20
11 50 50 50 84.00
12 50 50 100 88.99 0 0
100
13 50 0 50 84.58 100 80
80 60
14 0 50 50 84.52 60 40 1
Var 40 ble
15 0 0 100 88.84 iab 20 20 ria
50 50 50
le 2 0 Va
16 81.62 0

Figure 2: (Color online) This experimental design for a stain-removal index (SRI) for a coffee stain is
characterized by SRI coefficients (left) and can be visualized as response-surface models (right).
Esquejo et al.: Laundry Portfolio Optimization at P&G
Interfaces 45(5), pp. 444–461, © 2015 INFORMS 447

the right of the table is a graphical representation of Stain before wash Stain after wash
the treatments, with the shading corresponding to the
value of the stain-removal index (SRI) for coffee. In
this example, higher SRI values (darker shading) are
more desirable.
Our experimental designs were based on i-optimal
criteria; such designs minimize average variance of
prediction over the region of experimentation (Goos
and Jones 2011, Johnson et al. 2011). They also in-
cluded 16 variables with all two-way and selected
three-way interactions, producing third-order designs
Figure 4: (Color online) Stains are scanned before and after a wash exper-
with approximately 300 model terms. These variables iment; in this example, we use a coffee stain.
included all the key cleaning ingredients and wash
conditions of interest (e.g., surfactants and wash tem-
models. We accomplished model selection for each
perature). We used this design procedure for all stain
and whiteness models (approximately 60 responses). response using three stepwise regression techniques:
We generated empirical data for the design by P -value threshold, minimum-corrected Akaike infor-
making laboratory-scale formula prototypes, which mation criterion, and minimum Bayesian information
we physically tested in standardized wash proto- criterion; see Burnham and Anderson (2002, 2004),
Interfaces 2015.45:444-461.

cols. The stain removal and whiteness procedures we and Miller (1990) for a description of using Akaike
used were similar to ASTM method D4265-14 (ASTM and Bayesian information criteria for model selec-
International 2014), which involves creating standard tion. We used multivariate regression to quantify the
stain sets and characterizing their color before and selected models and validation metrics to determine
after wash (E) using image analysis. Figure 3 illus- the best model for each response, and we conducted
trates this procedure for stains for which measures are several levels of validation for each model to char-
assigned to each of several standard technical stains acterize prediction quality. One of the common tech-
that are washed together in a single experiment. Fig- niques involved quantifying standard fitting diagnos-
ure 4 shows an example of a standard coffee stain that tics for the data set used for model creation. These
has been processed with a given product test mixture metrics include R square, R square adjusted, root
(before and after wash). mean square error, lack-of-fit p-values, and other sim-
With the experiments in the design completed, ilar metrics. R square adjusted for the models ranged
our next step was to evaluate and select response from 0.55 to 0.95 with an average of 0.84. We used
models with R square adjusted below 0.70 only if
Before wash color – After wash color
Stain removal = –
Before wash color
our technology experts agreed that trends that the
model displayed were acceptable for business pur-
poses. All models in this design were deemed accept-
able. Figure 5 shows an example of these diagnostics
for coffee SRI.
Before After
wash Semi-empirical models are based, to some extent,
wash
on physical relationships. In this paper, we define
Temperature = t semi-empirical models as functional forms derived
Hardness = h
Soil level = s
from known equations (typically based on physical
laws and theorems), whose coefficients were deter-
Washer
mined by fitting the equation to a set of experimental
data. Finished-product density and intermediate den-
Figure 3: (Color online) A standard wash protocol is used when testing the sity were the primary semi-empirical models used;
stain-removal effectiveness of a mixture. both are nonconvex functions of mixture-ingredient
Esquejo et al.: Laundry Portfolio Optimization at P&G
448 Interfaces 45(5), pp. 444–461, © 2015 INFORMS

Actual by predicted plot is based on the material balance of water in the drying
 process.

Optimization
INSTANT ACTUAL

 The predictive models described offer the capability to


32)COFFEE

 determine cleaning and density properties of a given


 arbitrary mixture of ingredients without resorting to
 expensive and time-consuming physical prototypes.
 The next step was to determine ideal mixture ingredi-
 ent compositions to ultimately bring P&G products to
        the market while meeting stringent quality and manu-
32)COFFEEINSTANT facturing requirements. The optimization problem we
PREDICTED
describe determines the most economical composi-
PRSQuare
2-3% tion of mixtures, evaporation levels, and intermediate-
to-product assignments. In this section, we describe
Summary of fit the optimization problem at a high level to present a
basic statement of the problem, highlight complicat-
RSQUaRE 
ing features, and motivate the solution-methodology
RSQUaREADJ 
discussion. We include a more detailed mathematical
Interfaces 2015.45:444-461.

2OOTMEANSQUAREERROR 
formulation in the appendix.
-EANOFRESPONSE 
Intermediate batches and final products consist of
/BSERVATIONSORSUMWGTS 
mixtures of ingredients. Although ingredient propor-
tions determine the properties of these mixtures, the
Analysis of variance
manufacturing process does not simply blend pure
Sum of ingredients. Rather, premixes (mixtures of a small
Source DF squares Mean square F ratio
subset of ingredients) are sourced and mixed to
-ODEL      achieve a final-ingredient mixture. The eligibility of
%RROR     Prob > F premixes to be added to an intermediate batch versus
#TOTAL    * a final product is determined primarily by the nature
of the premix. Premixes with high water content are
Lack of fit typically added to intermediate batches so that the
Sum of F ratio water can evaporate. Some premixes (e.g., perfumes)
Source DF squares Mean square  can be unique to one product or available to all prod-
,ACKOFFIT     Prob > F ucts, either in the intermediate batch, postevaporation
0UREERROR     stage, or both.
4OTALERROR    Max r sq Figure 6 illustrates the analytical representation
 of material flowing through various stages of the
manufacturing process for a simple two-intermediate,
Figure 5: (Color online) Fit diagnostics, shown for the SRI response func- three-product example. Individual premixes may be
tion for a coffee stain, show strong predictive model performance and are eligible to go into intermediate mixtures, directly into
typical of stain results.
the final-product mixtures (as additives), or either. For
example, the figure shows that premix Pre1 can be
proportions. The appendix includes detailed func- assigned only to intermediates (B1 and B2), whereas
tional forms of these models. premix Pre3 can be assigned to intermediates and
Finally, we used a first-principles model of evapo- product Prod2. The additive Pre5 can be assigned
rative load for any given formulation to estimate the only to Prod3 directly.
impact on evaporative rate, an important manufactur- Percentages by weight of premixes used in total
ing consideration for detergent powders. This model production of intermediate batches (B1 and B2) and
Esquejo et al.: Laundry Portfolio Optimization at P&G
Interfaces 45(5), pp. 444–461, © 2015 INFORMS 449

Common intermediates In addition to these requirements, the following


Premixes e1
constraints must hold:
B1 (pre) B1 (post) 1. Ingredient and premix mass fractions must add
Pre1 up to 100 percent for all mixtures in the process.
e2 Products 2. Water content in the pre- and postevaporation
B2 (pre) B2 (post)
Pre2
intermediate mixtures must lie within a given range
Prod1
(typically between one percent and three percent mass
content for postevaporation).
Pre3 3. Each ingredient’s mass fractions must lie within
Prod2
predefined bounds to ensure the integrity of the phys-
Pre4 ical models.
Prod3
4. Finished products must have density values
within predefined bounds.
Pre5
5. Finished product SRI and whiteness values must
be bounded below by benchmark values.
6. A predefined number of intermediates must be
used.
Figure 6: The layout defines the various decision variables and constraints Items 1–3 are linear constraints and do not impose
of the optimization problem.
undue computational burden on the solution pro-
Interfaces 2015.45:444-461.

cess; however, items 4 and 5 enforce density, SRI, and


products (Prod1, Prod2, and Prod3) are determined
whiteness constraints, introducing additional nonlin-
as part of the optimization problem. In turn, these
ear, nonconvex constraints to the formulation. Product
decision variables determine mass percentages of raw
density is expressed in terms of a rational function,
materials in the total production of intermediate and
and SRI and whiteness are third-order polynomial
product mixtures, as determined by the given premix
functions of the mixture-composition variables. In
compositions. Because intermediates must adhere to
combination with the previously described bilinear
stringent water-content requirements, an evaporation
term and binary decision variables, these features
phase that produces a postevaporation mixture is nec-
make this a difficult mixed-integer nonlinear program-
essary, as the distinction between pre- and postevap-
ming (MINLP) problem.
orated intermediates in Figure 6 indicates.
The optimization objective is to minimize the total
The extent of evaporation of intermediate batches
cost of premixes used in the mixing process, weighted
is a control variable of the manufacturing process and
by product dosage per stat unit (i.e., a unit of de-
must be prescribed as a decision variable for each
mand) and product production volume targets. This
intermediate batch; because it is accomplished using
objective represents the total cost of production. The
spray drying technology, it is also subject to maxi-
appendix provides a detailed mathematical formula-
mum evaporation constraints based on plant restric-
tion of the MINLP, including all sets, decision vari-
tions. The appendix provides mathematical details
ables, constraints, and objectives.
about the role of this variable; in this discussion, we
simply point out that it introduces into the formula-
tion a difficult bilinear term that cannot be eliminated. Optimization Solution Methodology
Each product can derive a portion of its mixture As previously stated, a requirement of the problem
from at most one postevaporated intermediate. This is that a given number of intermediate batches must
requirement permits a continuous manufacturing pro- be used in the mixing process to produce a set of
cess flow, which avoids capital investment for stor- products, where each product can take a portion of
age and handling of the intermediate batch; however, its composition from exactly one intermediate. Any
it limits the ability to channel various intermediate number of products can use an intermediate. We can
batches to each product. The intermediate-to-product interpret this as a set-partitioning problem with non-
assignment is a binary decision variable. linear side constraints; that is, we aim to partition
Esquejo et al.: Laundry Portfolio Optimization at P&G
450 Interfaces 45(5), pp. 444–461, © 2015 INFORMS

the set of products into subsets, such that an inter- implement a new work process. The use of optimiza-
mediate is assigned to each subset of products that tion tools requires a set of skills best defined by an
must satisfy various nonlinear constraints, such as optimization triad (Figure 7) that includes functional,
performance requirements. This interpretation of the data, and optimization experts.
problem is closely related to the pooling problem Functional experts are typically a small number of
(Bodington and Baker 1990), and can be shown to individuals from different R&D functions, including
be an NP-hard mixed-integer nonlinear (nonconvex) consumer, formulation, and process. Data experts are
program. one or two individuals who have access to all the
Previous approaches to solving pooling problems necessary information, such as material pricing and
include relaxation and discretization strategies (Gupte material balances, and are typically skilled in visual
et al. 2013), Benders decomposition (Floudas and analytics tools, such as JMP statistical tools. Optimiza-
Aggarwal 1990), Lagrangian relaxation (Visweswaran tion experts are staff members who have the neces-
and Floudas 1990), branch and cut (Audet et al. 2004), sary programming skills to interact with the optimiza-
and mixed integer linear programming (MILP) (Dey tion models at the SAS code level. Figure 7 defines
and Gupte 2013). These approaches do not directly the responsibilities of expert in each category.
apply to the variant we consider in this paper. Most of We defined a new work process (Figure 8), which
the effort to date has focused on addressing the bilin- has enabled this multifunctional, multiskilled team
ear terms in the problem; however, the evaporation to efficiently use the new capability; we describe the
Interfaces 2015.45:444-461.

process, the requirement that assignments must be components as follows.


binary decisions, and nonconvex performance models 1. Problem definition: Functional experts frame the
violate basic assumptions of much of the work con- problem to be solved and the research questions to be
ducted to date. addressed. For example, we researched and analyzed
State-of-the-art solution methodologies for the pool- processes that would (1) allow P&G to deliver signif-
ing problem that attempt to prove optimality are typ- icantly better product performance at equal or lower
ically restricted to problems of (in equivalent terms) a cost over its current processes, and (2) reduce the num-
few dozen ingredients, premixes, intermediates, and ber of intermediates while maintaining performance.
products; Dey and Gupte (2013) provide example 2. Knowledge development: Functional experts
instance sizes, where the number of inputs and out- start building the problem’s framework by collecting
puts is less than 100. Several features of the prob- knowledge and building a common state of under-
lem we address in this paper make it a significantly standing across the team. We define knowledge as
harder generalization of the classical pooling prob- existing models, heuristic approaches, and the simpli-
lem, although its size is comparable to instances fication of assumptions, all of which must form a basis
addressed by the best-known methods. Furthermore, for agreement. For example, if different density mod-
the best-known methods often take many hours to els exist, the team must come to a consensus on which
reach a small optimality-bound gap. For our pur-
Functional expert Optimization expert
poses, solutions must be produced in an order of min-
$EFINESPROBLEMTO $EFINESOPTIMIZATION
utes (to allow scenario exploration, which is critical to BESOLVED CREATES FRAMEWORK WRITES
P&G formulation practices); therefore, we have taken APPROPRIATEMODELS CODE RUNSSCENARIOS
a heuristic approach to solve this optimization prob- DEFINESCONSTRAINTS
lem. The appendix provides a detailed explanation of
the optimization solution methodology.
Data expert
'ATHERSAPPROPRIATEINPUTDATAINTO
Implementation and Usage DESIREDFORMAT ANALYZESSOLUTIONS
USINGVISUALANALYTICSTOOLS
Team and Workflow
To best utilize this new analytical capability, P&G had Figure 7: (Color online) This figure shows the optimization triad that P&G
to devise a new way of organizing our team and also adopted as part of its standard process.
Esquejo et al.: Laundry Portfolio Optimization at P&G
Interfaces 45(5), pp. 444–461, © 2015 INFORMS 451

7ORKPROCESS

!SSUMPTION
0ROBLEMDEFINITION +NOWLEDGE -ODELDEVELOPMENT
FUNCTIONALEXPERT DEVELOPMENT FUNCTIONALEXPERT
FUNCTIONALEXPERT

/PTIMIZATION 0ORTFOLIODATA
7RITEOPTIMIZATION
PROBLEMFORMULATION GATHERING EG VOL
CODE
OPTIMIZATIONEXPERT COST FORMULATION
OPTIMIZATIONEXPERT
DATAEXPERT

)TERATIVETESTING /PTIMIZED
-ATHEMATICAL #HURN!NALYSIS
RECOMMENDATION
VALIDATION !LL
FUNCTIONALEXPERT
OPTIMIZATIONEXPERT
Interfaces 2015.45:444-461.

Figure 8: (Color online) The optimization work process at P&G follows a sequence of nine interrelated steps.

density model to use. Additional data or analysis may new research question, constantly verifying assump-
be required if the team cannot come to an agreement tions and validating preliminary results with other
because of missing or conflicting information. experts and stakeholders.
3. Model development: Functional experts develop 7. Iterative testing and mathematical validation:
models for new-product parameters or adapt (e.g., Optimization experts participate with other members
linearize) current models to make them more suitable of the team in a cycle of iterative testing to mathemat-
for optimization. ically validate models, ensuring that constraints are
4. Portfolio data gathering: Data experts collect and correctly interpreted and observed and that optimum
format all the information needed for the framework. solutions are robust. In this step, parameters (e.g.,
This proved to be a challenging step because we the number of multistart points and stopping criteria)
encountered multiple databases with missing rela- are tuned, and model infeasibilities, often caused by
tional keys. The process can be quite manual in some products with too-stringent performance constraints,
cases. are commonly found. At this stage, steps must be
5. Optimization problem formulation: Optimiza- taken to ensure feasibility; revisiting the constraint
tion experts interpret the research question(s) and, bounds is an example.
using the knowledge collected, models, and portfolio 8. Churn and analysis: In this step, the entire team
information as context, create or adapt the mathemat- exercises the optimization engine to optimize multiple
ical framework for optimization by defining or mod- scenarios to answer the research question(s). Scenarios
ifying variables, objective functions, and constraints. can include changes in performance requirements, the
Optimization experts start the modeling work as soon number of intermediates allowed, constraint bounds,
as the research question has been defined, and further materials allowed to be used (and where they can be
refine the optimization models as additional knowl- used), or a combination of these scenarios. For highly
edge is generated and the portfolio data are compiled. complex research questions, P&G implements a churn
6. Write optimization code: Optimization experts event; in such an event, all members of the team are
write or adapt SAS/OR code as needed to address the colocated for two to four days. They focus all their
Esquejo et al.: Laundry Portfolio Optimization at P&G
452 Interfaces 45(5), pp. 444–461, © 2015 INFORMS

time on a common set of problems, and their objec- stopping criterion at three intermediates at an esti-
tive is to produce data to inform decision making. mated gap of three percent for all optimization runs.
During churn, the team occupies a common room Figure 9 shows solution objective values (total cost
equipped with physical and digital visualization tools difference for annual production) of a typical optimi-
that aid the work process. Poster-sized printouts dis- zation, which we represent as the difference between
play a master list of all scenarios, and key parame- our solution output objective and the annual cost of
ters are captured and color coded to differentiate the a historical production run. An obvious feature of the
scenarios. They analyze completed scenarios on a dis- problem we consider here is that as the number of
play that consists of eight 42" high-definition televi- intermediate batches increases, the globally optimal
sion screens configured to behave as a single monitor, objective must not increase, because the model allows
allowing high-resolution visualizations to be spread more flexibility in tuning each product’s assigned
across a large area. The team typically reviews the intermediate. The figure reflects this, and we note that
results using JMP software, which permits interactive the algorithm’s flow was motivated to ensure that this
visualization and analysis. check to verify rationality would never be violated
9. Optimized recommendation: Functional experts despite the presence of nonconvexity.
take results from the analysis and formulate a recom- Reporting such a multiple intermediate solution
mendation. Recommendations can be as simple as a proved valuable to management because an immedi-
new set of formulations to meet a new requirement, ate comparison of neighboring solutions could show
or as complex as a multistage strategy for a portfolio
Interfaces 2015.45:444-461.

the benefits of increasing (or decreasing) the number


of products, which might include the introduction of of intermediate batches in production—a costly, but
new technologies. sometimes beneficial, manufacturing investment. For
In the course of this project, we learned that a mind- example, the marginal benefits of increasing the num-
set shift is required to solve for the entire portfolio ber of intermediate batches from 14 to 17 are minimal;
rather than for one product at a time. Traditionally, only at 18 intermediates does a significant change
projects have addressed formulation changes for one occur (because of the discrete nature of the problem),
or only a small subset of products, while preserv- which may justify the added complexity and cost.
ing the composition of most products in the portfo-
lio. Once this mind-set shift happened, however, we
were able to identify (and test) portfolio-management /PTIMALCOSTSDIFFERENCEVSBENCHMARK
  
strategies, which led to some unique options that
   3OLUTION
P&G had not previously uncovered. "ENCHMARK
#OSTDIFFERENCEVSBENCHMARK

  

Outputs and Results   

Because the algorithm we present in this paper is   

based on building an approximation of the power set   


of products, each optimization can be easily extended   
to solve for various numbers of intermediate batches 
by solving the selection step for as many indepen- n   
dent subproblems as there are products—from one n   
intermediate for all products to the singleton cases. n   
We exploited this aspect of the algorithm to pro- n   
vide a range of intermediate batch configurations for                      

each optimization run for marginal additional run .OOFBATCHES

time, running the set-covering subproblems in par-


allel. When running in this mode, one must deter- Figure 9: (Color online) Multiple intermediate-solution annual costs show
a potential reduction of $4 million when the number of intermediate
mine where to apply the stopping criterion. Because batches is held constant for a 14-batch instance. An alternative interpre-
P&G’s production focus is on number of interme- tation suggests that, at similar costs and levels of quality, production can
diates greater than or equal to four, we apply our occur with only six intermediate batches.
Esquejo et al.: Laundry Portfolio Optimization at P&G
Interfaces 45(5), pp. 444–461, © 2015 INFORMS 453

As we mention previously, each optimization in- Cost diff. vs. bench


Intermediates ($ in millions) Bounds (%) Run time (sec)
stance was referenced against a related historical
benchmark production run (created without the ben-
3 −1508 2035 45805
efits of an analytical approach). Our expectation for 4 −1609 1077 25300
the success of this project was that the solution pro- 5 −1800 1019 11002
vided by the framework would always improve on 6 −1808 0078 11004
7 −1902 0058 11005
the benchmark’s premix annual costs. This is a rea-
8 −1904 0045 11006
sonable expectation because, for a given number of 9 −1906 0036 11006
intermediate batches, the benchmark is typically a 10 −1907 0029 11007
11 −1908 0022 11009
feasible mixture. Comparing the 14-intermediate solu-
12 −2000 0015 11100
tion annual cost with the benchmark value in Figure 9 13 −2001 0009 11101
shows that this requirement was met in this exam- 14 −2002 0004 11102
ple, and that the solution typically produces savings 21 −2003 0000 3206

in the order of magnitude shown (between $0.5 and Table 2: The table shows results for instance “3bii.” The bolded values
$6 million). represent a best comparison to the benchmark solution, which has 12
Few instances currently exist at P&G because sta- intermediate batches.
tistical and semi-empirical models have only recently
been developed for entire product portfolios. Table 1
Interfaces 2015.45:444-461.

lists two important powder instances we used in bolded row represent a direct comparison with the
this work and also shows a number of ingredi- benchmark intermediate configuration.
ents, premixes, products, response-surface models, The most direct application of our work has been in
and benchmark intermediates that represent typical P&G’s North American dry laundry portfolio, which
problem sizes. consists of more than 20 unique formulations. Our
The optimization tool produces significantly bet- objective was to study different strategies for sim-
ter results than the benchmark for both instances in plification and savings, while delivering the same or
the table; its run times are very reasonable, given the better consumer-relevant cleaning performance. The
needs of the P&G work process. Tables 2 and 3 sum- churn team spent two days running and analyzing
marize the results for these instances, listing most
intermediate configurations from three to the num-
ber of products in the instance. The tables show cost Cost diff. vs. bench
differences from the benchmark (in millions) and run Intermediates ($ in millions) Bounds (%) Run time (sec)
times at which each intermediate configuration satis-
fied its stopping criterion. The longest run time is the 3 −7604 2046 25606
4 −7908 1042 25607
total run time of the process. Note the cost improve-
5 −8109 0079 25608
ments when compared to the benchmark even when 6 −8206 0058 25700
running at the minimum number of intermediate 7 −8300 0043 25702
batches (three) for both instances. The values in the 8 −8305 0028 25703
9 −8307 0021 25704
10 −8309 0015 25705
11 −8400 0012 25707
12 −8401 0009 25708
Instance name Ingredients Premixes Products RS Intermediates
13 −8402 0007 25800
14 −8403 0005 25801
3bii 54 38 21 50 12 15 −8403 0002 25803
Wenlock 42 34 25 38 13 16 −8404 0001 25804
25 −8404 0000 6805
Table 1: This table summarizes the size of two representative instances
in P&G’s portfolio. The term RS refers to the number of response-surface Table 3: In these results for the Wenlock instance, the bolded values rep-
models, and the Intermediates column represents the number of interme- resent a best comparison to the benchmark solution, which has 13 inter-
diate batches in the benchmark solution. mediate batches.
Esquejo et al.: Laundry Portfolio Optimization at P&G
454 Interfaces 45(5), pp. 444–461, © 2015 INFORMS

more than 30 scenarios. Here, we discuss two exam- • Redefining the cleaning vision: Differentiating
ples from the study results. performance between brands and maintaining brand
• Increasing the number of intermediates allowed equity.
from five to 10 would result in additional formula cost • Exploring out-of-the-box concepts: Exploring cur-
savings. Compared to a five-intermediate benchmark, rent production constraints and raw material ingredi-
using the optimization procedure would result in ents, but with the flexibility to integrate deviations
annual cost improvements of $2 million for five that may lead to better results in performance and
intermediates and $5 million for 10 intermediates. process.
Although a consequence of this strategy is to increase • Managing what-if scenarios: Saving time and
complexity at the manufacturing site for handling resources.
a higher number of intermediate batches, this is a • Multifunctional integration: Performing a more
justifiable decision based on this demonstrable cost robust proposition versus isolated optimization ef-
reduction. forts based on function.
• The introduction of a new active ingredient (cur-
rently not in the North American powder formula- Summary of Benefits
tion, but available in empirical and semi-empirical Portfolio optimization is changing the way we do
models) across the whole portfolio would generate product development at P&G. Previously, we were
annual savings in excess of $20 million, while deliv- often limited to one of two strategies; we developed
Interfaces 2015.45:444-461.

ering target performance profile (i.e., cleaning of dif- each product individually, resulting in highly complex
ferent stains) for each product. P&G has incorporated portfolios that required high numbers of intermediate
this knowledge into its short- to mid-term strategies batches, or we imposed simplification strategies that
for this part of our business. resulted in higher formulation costs.
We have estimated that without the portfolio opti- Portfolio optimization allows us to test formulation
mization tool, twice as many staff members would and simplification strategies against the whole port-
have to work for far longer periods of time to run folio, giving us a realistic estimate of the potential
30 scenarios on 20 formulations using the former impact of these strategies. Fast iteration cycles allow
us to evaluate multiple strategies in a short time, dis-
single-product-at-a-time approach and would pro-
carding elements that will bring little value and com-
duce inferior results.
bining elements that provide significant advantages.
User sponsorship and adoption have been instru-
The ability to analyze an entire portfolio simultane-
mental to the success of this project. In an email to
ously is also changing the way product designers
the team, Christian Becerra, P&G senior researcher
think about performance, because we can now more
and lead formulator for the North American powder
easily differentiate performance among the products.
business, provided a set of additional benefits of our
Thus, we can make smarter decisions about formu-
portfolio optimization framework that go beyond the lation and simplification strategies and respond with
savings described earlier (Becerra 2014): agility when needed.
• Smart optimization: Identifying a formula and As lead strategies are identified and the formula-
process strategy that meets our criteria, removing tion for the full portfolio is generated, some phys-
chemistry that will not deliver the desired perfor- ical testing is required to confirm that the required
mance profile to the consumer. This leads to smart performance and physical properties of the products
savings. are indeed met. This is especially true for solutions
• Next level of optimization: Going beyond the tra- that are near the minimum or maximum values of
ditional single-formula to full-portfolio optimization the input ranges, because the confidence intervals of
(i.e., the ability to see the big picture). the predictions are typically at their widest in these
• Flexibility: Integrating technical features and an ranges. Future phases of this project will address this
understanding of consumer needs to make more need, and continuously expand and improve the qual-
robust portfolio propositions. ity of predictive and optimization models.
Esquejo et al.: Laundry Portfolio Optimization at P&G
Interfaces 45(5), pp. 444–461, © 2015 INFORMS 455

As a direct result of using this optimization plat- Product Density


form and measuring its value to our business, we
1
are now making the following enhancements to our FDk 4zk 1 xk 5 = Pn 0
0
4z
i=1 i /4 i i + xk /CDb 4‡b 5
f 55
processes:
• Evolving P&G and its work processes to better zk : Vector of mass fractions zik of finishing additive i in
use this and other operations research capabilities to finished product k.
xk : Mass fraction of assigned intermediate in production of
define and execute the best portfolio strategies.
finished product k.
• Investing in resources to improve our data i : Density of finishing additive i.
management system and automate data pull and fi : Packing factor of finishing additive i.
formatting.
• Building more models into the optimization frame- Optimization Model
work and continuing to do so as new models become Notation
available, potentially including consumer models and
first-principles chemistry models. Sets
• P: Set of premixes.
Finally, although the benefits of this tool have been • I: Set of ingredients (raw materials).
well demonstrated in our laundry-powder business, • B: Set of intermediate batches.
many reapplication opportunities remain at P&G. In • K: Set of finished products.
the coming years, we expect to continue to develop
Parameters
Interfaces 2015.45:444-461.

this capability further for our laundry business and


create similar capabilities for other P&G businesses. • n: Number of intermediates required in the solution.
• v̄pb : Mass percentage upper bound of premix p in pro-
duction of pre-dried intermediate b.
Appendix 0
• v̄pb : Mass percentage upper bound of premix p in pro-
Semi-empirical Models duction of postevaporated intermediate b.
Stain-removal performance is predicted by the stain- • z̄pk : Mass percentage upper bound of premix p in pro-
removal index (SRI) response function, which has the fol- duction of finished product k.
lowing form: • ip : Mass percentage of ingredient i in production of
premix p.
SRI = C0 + C1 v1 + C2 v2 + C3 v1 v2 + · · · 1 • cp : Cost of premix p, in $ per metric ton.
• qks (stat factor): Number of doses in a stat unit for prod-
where Ci represent coefficients and vi represent design vari-
uct k.
ables: wash concentrations (milligrams per liter) of chemical
ingredients and wash conditions (e.g., temperature). White- • qkv : Production target of product k: number of stat units
ness models have a similar form. sold per year.
• qkw (wash volume): Volume of water in a single wash
Intermediate Batch Density for product k, in liters.
CDb 4‡b0 5 = C0 CIb 4‡b0 5 + C1 V3 + C2 V4 + · · · • qkd (“dosage”): Grams of product k in wash machine
for one wash cycle, in liters.
CIb : “True density” of an intermediate batch, also • R̄k : Maximum rate of evaporation limit for product k.
known as absolute density; the density of only the • r¯b : Maximum evaporation capability for intermediate
solid components of the batch post-evaporation. batch b.
CDb : “Bulk density” or the density accounting for the
mass of solid components and air entrapped in the Decision Variables
particle. • vpb ∈ 601 v̄pb 7: Mass percentage of premix p in produc-
‡b0 : Vector ‡ib 0
of postevaporation mass fraction of tion pre-evaporated intermediate b.
ingredient i in intermediate b. • eb : Evaporation variable for intermediate b.
C0 1 C1 1 C2 1 0 0 0 : Regression coefficients. • zpk ∈ 601 z̄pk 7: Mass percentage of premix p in produc-
Vj : Variables that define processing and chemistry tion of finished product k as an additive (not subject to
parameters (e.g., temperature, hardness, ‡b0 ) evaporation).
1 • xk ∈ 601 17: Mass percentage of intermediate component
CIb 4‡b0 5 = Pn 0 in production of product k.
i=1 ‡ib /i
• ybk = 1 if finished product k is assigned to intermediate
i : Liquid density of pure component i. batch b, 0 otherwise.
Esquejo et al.: Laundry Portfolio Optimization at P&G
456 Interfaces 45(5), pp. 444–461, © 2015 INFORMS

Because mixing decisions are made at a premix level, Empirical models impose constraints on product perfor-
variables v, z, and y take a p ∈ P index, thus specifying pre- mance. The term Fk 4”k 5 represents a vector for each product
mix composition. The evaporation process is expressed in of all SRI and whiteness functions, each row characterized
terms of ingredient compositions, requiring us to calculate by third-order polynomial expressions of the product’s com-
ingredient mass percentage quantities pre- and postevapo- position ”ik and parameters temperature, hardness, and soil
ration for each ingredient i in each intermediate b: level. Products created by the optimization must achieve a
X minimum level of performance, which is defined by vector
‡ib = ip vpb 1 ∀ i ∈ I1 b ∈ B1 (1) fk for each product:
p∈P
0
‡ib = eb ‡ib 1 ∀ i ∈ I\8w91 b ∈ B1 (2) Fk 4”k 5 ≥ fk 1 ∀ k ∈ K0 (10)
0
eb 41 − ‡wb 5 = 1 − ‡wb 1 ∀ b ∈ B1 (3) Semi-empirical models impose constraints on product
density,
where the subscript w is used to denote water. Similarly,
empirical and semi-empirical functions are expressed in F Dk ≤ F Dk 4zk 1 xk 1 ybk 5 ≤ F Dk 1 ∀ k ∈ K1 (11)
terms of ingredient mass percentage at the product level, where we recall that F Dk 4zk 1 xk 1 ybk 5 is a nonlinear function
requiring us to calculate these values for each ingredient i representing the density of product k, which is dependent
in each product k: on ybk through its corresponding intermediate batch den-
X X 0 sity CDb .
”ik = ip zpk + xk ybk ‡ib 1 ∀ i ∈ I1 k ∈ K0 (4)
p∈P b∈B
Finally, we recall that a requirement of the process is that
each product is assigned to exactly one intermediate batch:
Constraints. Fundamental physical requirements that
Interfaces 2015.45:444-461.

X
characterize the mixing process are modeled as constraints. ybk = 11 ∀ k ∈ K0 (12)
b∈B
Intermediate and final-product mixtures have mass percent-
age ingredient contributions that must add to 100 percent: Objective. The optimization objective is to minimize the
X X total cost of premixes used in the mixing process, weighted
vpb = 11 ∀ b ∈ B and zpk + xk = 11 ∀ k ∈ K0 (5) by product dosage per stat unit (qks , a unit of demand) and
p∈P p∈P
product production-volume targets (qkv ), using at most n
The rate of evaporation Rk is limited by intermediates:
 
X s vX X
Rk = qkv xk 0 min qk qk cp zpk + xk ybk eb vpb 0 (13)
X
4‡w1 b − ‡w1 b 5ybk ≤ R̄k 1 ∀ k ∈ K0 (6)
b∈B k∈K p∈P b∈B

Additionally, there are physical limitations on the amount A useful quantity in the presentation to follow is the cost
of water that can be evaporated in the intermediate batch: of production of a subset of products A ⊆ K for a given set
of values x, y, e, v, and z:
0
‡w1 b − ‡w1 b ≤ r¯b 1 ∀ b ∈ B0 (7) X s vX

X

cA = qk qk cp zpk + xk ybk eb vpb 0
Empirical and semi-empirical models are based on exper- k∈A p∈P b∈B
imental designs that are valid only within specific values,
and accuracy can degrade severely if extrapolated beyond Optimization Solution Methodology
these bounds. Furthermore, compositions cannot differ
Our algorithm is based on a column-generation heuris-
drastically from benchmark mixtures, and the water content
tic, where a set-covering master problem interacts with
of powder products must be strictly controlled. Therefore,
independent subproblems that prescribe intermediate-to-
ingredient mass percentage values must lie within prede-
product groupings and mixture compositions. The algo-
fined lower and upper bounds:
rithm is based on the following sequence of steps:
”ik ≤ ”ik ≤ ”¯ ik 1 ∀ i ∈ I1 k ∈ K0 (8) 1. Singleton
2. Grouping
Similarly, bounds must be imposed on ingredient composi- 3. Configuration
tions in the intermediate batches: 4. Selection
5. Return to step 2 until convergence.
0 0 0
‡ib ≤ ‡ib ≤ ‡¯ ib 1 ‡ib ≤ ‡ib ≤ ‡¯ ib 1 ∀ i ∈ I1 b ∈ B1 (9)
Singleton Step
which include, most importantly for the role they play in To start the process, we solve for the artificial case in
the manufacturing of intermediate batches, constraints on which each product is allowed to have its own dedi-
water content. cated intermediate batch. This is equivalent to specifying K
Esquejo et al.: Laundry Portfolio Optimization at P&G
Interfaces 45(5), pp. 444–461, © 2015 INFORMS 457

Premixes SRI requirements for the same stains often share an inter-
mediate to reduce costs; in isolation, they would have very
Pre1 similar intermediate batches.
Common intermediates To exploit this observation, we define a metric of simi-
larity between optimal (or best-known) intermediate com-
Pre2 e1 positions of singleton solutions. For any two singletons k
B1 (pre) B1 (post)
Products and l,   
kl = cp pk − pl  (14)
Pre3 Prod1 p∈

represents a sum of the absolute-value difference of mass


percentage of premixes in their respective singleton poste-
Pre4
vaporation intermediates, weighted by the cost of premix,
where we have used the product indices to represent corre-
sponding singleton pairs.
Figure A.1: The singleton subproblem layout for Prod1 is much sim- The goal of the grouping step is to generate candidate
pler than the full-problem layout of Figure 6 and is independent of other groups of products that minimize the sum of kl , such that n
products. intermediate batches are used. We accomplish this by defin-
ing a model based on bipartite assignments, where each
product is eligible to be paired with each singleton’s optimal
independent problems, each with n = 1. For an example port- intermediate. Binary variables determine which intermedi-
folio P1 P2 P3 , this requires us to solve three indepen- ates should be used, and arc variables determine prospec-
dent mixture-optimization problems comprising P1 , P2 , tive intermediate-to-product pairings. These are then inter-
Interfaces 2015.45:444-461.

and P3 . Figure A.1 shows the corresponding Prod1 sin- preted to specify product groups that are appended to 
gleton of the problem illustrated in Figure 6. (each composed of a subset of products).
We use the SAS/OR interior point nonlinear program-
ming solver for the singleton subproblems, which are inde- Sets
pendent and can therefore be solved in parallel by enabling • : Current groupings pool.
the SAS cofor multithreaded processing capability. Compli- •  = k l ∈  ×  k ≥ l.
cations exist, however, primarily because of bilinear terms
Parameters
and nonconvex empirical and semi-empirical functions.
• kl : Similarity measure; see Equation (14).
We address these complications by employing the multi-
start mechanism of the SAS nonlinear programming (NLP) Variables
solver (also threaded), which aims to improve the likelihood • vl = 1 if the intermediate of product l is used in calcu-
of finding globally optimal solutions. Note that the single- lating kl for all k to be grouped with l, 0 otherwise.
ton step subproblem is a specialization of the configuration • ukl = 1 if product k is assigned to intermediate of sin-
step problem, where  =  and ykk = 1. gleton l, 0 otherwise.
Although global optimality is neither provable nor guar- • A = 1 if group A ∈  is selected into the solution, 0
anteed, we will describe a method for improving the ulti- otherwise.
mate quantity that is to be derived from these solutions: the
Grouping Problem Formulation
globally optimal cost ck of each singleton. For now, we begin

to build a groupings pool  as the union of all singletons. Minimize kl ukl (15)
In the previous example,  = P1  P2  P3 . k l∈

subject to ukl = 1 ∀ k ∈  (16)
Grouping Step l∈
We exploit a physical observation to generate a relatively 
small number of promising product groups to approximate vl = n (17)
l∈
this idea. The observation relies on the premise that expect-
ing products with similar performance requirements to be ukl ≤ vl ∀ k l ∈  (18)
able to benefit from extracting a portion of their mixture vk ≤ 1 − ukl ∀ k l ∈  k = l (19)
from the same intermediate is reasonable. Singletons are
ideal chemical compositions for products because they ben- ukl ∈ 0 1 ∀ k l ∈  (20)
efit from dedicated intermediates. Grouping finished prod- vl ∈ 0 1 ∀ l ∈  (21)
ucts based on the similarity of their singleton intermediate
batch compositions is known to be beneficial by observation Inequality (19) ensures that if a product has been assigned
in practice. For example, products that must meet stringent to an intermediate batch of another product’s singleton,
Esquejo et al.: Laundry Portfolio Optimization at P&G
458 Interfaces 45(5), pp. 444–461, © 2015 INFORMS

its own singleton intermediate batch becomes ineligible for


Premixes
grouping other products.
Groups implied by ukl are appended to  for further Pre1
steps in the algorithm (i.e., configuration and selection).
Common intermediate
Our definition of  ensures no symmetry in the prob-
lem, allowing us to use ukl to uniquely map  to . For Pre2 e
Pre Post
example,  
1 Products
 
u=1 0  Pre3
0 0 1 Prod2

has a one-to-one mapping to P1  P2  P3 . Multiple itera-


Pre4
tions of the algorithm build out  by appending nonzero Prod3
columns of each solution u to . For example,

 = P1  P2  P3  P1  P2  Pre5

created by appending the above solution to a groupings


pool that has been initialized with singletons, is represented Figure A.2: The configuration subproblem layout shows that only one com-
in matrix form as mon intermediate exists, eliminating the need for the binary variable that
  assigns products to intermediates.
1 1
Interfaces 2015.45:444-461.

 
= 1 1
1
This variant of the problem has no integer variables,
Because the purpose of this step is to enrich  to provide allowing us to use standard NLP solvers. Because the sin-
a better approximation of the optimal members of , gleton problem can be interpreted as a special case of the
each iteration (i.e., each time the grouping problem is called) configuration problem where only one product exists, we
must produce n groups of which at least one is not currently implement the singleton computation by using the configu-
in . We accomplish this by adding to the above formula- ration model code.
tion variable A and constraints In the configuration step, we solve the related NLPs to
identify (locally) minimum-cost values of producing these
 
1 − ulk  + ulk ≥ 1 − A ∀ A ∈  k ∈  (22) independent groups of mixtures. These problems are solved
l∈A l∈\A in parallel using SAS NLP solver-threaded capability.

and Selection Step



A ≤ n − 1 (23) The selection step is simply a cardinality-constrained set
A∈
covering of  using groups in  as eligible subsets. The
where we interpret A as columns of , l ∈ A as rows of  goal is to minimize the total optimal cost of production
with values of 1, and l ∈ \A as rows of  with values of 0. 

n=
We can interpret the variable as follows. For an existing c cAi
group A ∈ , if A = 0, then Inequality (22) ensures that A is ∗
Ai ∈T n
not a group in the solution of the grouping step as specified
∗ ∗
by u, and if A = 1, the constraint is relaxed. Inequality (23) with an optimal partition T n , where T n  = n. Because 
allows the model to relax this condition up to n − 1 times. is a relatively small set, we can easily solve this problem
For example, when solving for a five-intermediate problem, with the SAS/OR MILP solver.
Inequalities (22) and (23) together ensure that a new solu-
tion is produced that allows for at most four intermediate Parameters
batches already in . • xkA : Mass percentage of intermediate component in
production of product k for solution of group A.
Configuration Step • vpA : Mass percentage of premix p in production of pre-
Given a group of products, the production optimization evaporated intermediate for solution of group A.
problem becomes a continuous NLP. Figure A.2 illustrates • ckA : Cost of producing product k using group A
the example of Figure 6 when we impose a group Prod2, solution:

Prod3. Note that there is a single intermediate, and there- ckA = qks qkv cp zpk + xkA eA vpA  (24)
fore no need to identify the intermediate batch by index. p∈
Esquejo et al.: Laundry Portfolio Optimization at P&G
Interfaces 45(5), pp. 444–461, © 2015 INFORMS 459

Variables to a known solution of A because of the effect of noncon-


• sA = 1 if group A is used in the solution, 0 otherwise. vexity. The process ensures that we provide the best-known
• wkA = 1 if intermediate of product k is derived from coefficients for the objective function in the selection step by
group A, 0 otherwise. exploiting the hierarchy in the problem, even if we have not
converged to global optimality for every NLP calculation in
Selection Problem Formulation the configuration step.
XX We further exploit this idea to augment A with all parti-
Minimize ckA wkA (25)
A∈A k∈A
tions of its members, because initializing all such partitions
X with the solution of their supersets is possible. For example,
subject to wkA = 1 ∀ k ∈ K1 (26) 
A∈A2 k∈A A = 8P1 918P2 918P3 918P4 918P5 918P2 1P3 918P1 1P4 1P5 9 (33)
X
sA = n1 (27) can be augmented with 88P1 1 P4 91 8P1 1 P5 91 8P4 1 P5 99, with
A∈A
X each new subgroup being initialized with the restricted por-
wkA ≥ sA ∀ A ∈ A1 (28) tion of s8P1 1 P4 1 P5 9 and c8P1 1 P4 1 P5 9 .
k∈A We cannot guarantee global optimality for the configu-
wkA ≤ sA ∀ A ∈ A1 k ∈ A1 (29) ration step because of the nonconvexity of the problem;
therefore, cA can only be regarded as an upper bound to
sA ∈ 801 19 ∀ A ∈ A1 (30) the production cost of a group A. Ideally, for the solution
wkA ∈ 801 19 ∀ A ∈ A1 k ∈ A0 (31) methodology to converge with some confidence of optimal-
ity, it would be helpful to identify lower bounds cA for costs
Algorithm Enhancements (Augmentation and Bounding) of groups in P4A5 for which we have not solved the associ-
Interfaces 2015.45:444-461.

Initially, when we start with singletons and the first round ated NLPs. We illustrate this using the following example.
of groupings, A contains at most K + n subsets. We note Consider a scenario in which we are solving a five-
that the similarity-based grouping step attempts to heuristi- product, two-intermediate problem and have built A de-
cally exploit an observation of optimal intermediate compo- fined in Equation (33). Such a pool would be constructed
sitions but is not guaranteed to lead to optimal intermediate by the solution of the singleton step with one augmentation
batch groups. Furthermore, because of the nonconvexity of from the grouping step that produces 88P2 1 P3 91 8P1 1 P4 1 P5 99.
the problem, the configuration step likely produces solu- We solve an NLP for each member of A to calculate its cor-
tions that are not globally optimal. responding cost. Although we have not solved the NLP for
To address these issues, we iterate through the group- the group 8P1 1 P2 1 P3 9, we can estimate its lower bound as a
consequence of Inequality (32),
ing, configuration, and selection steps to accomplish two
goals: (1) improve our approximation of the optimal region 
c8P1 1 P2 1 P3 9 2= max 4c8P1 9 + c8P2 9 + c8P3 9 51 4c8P1 9 + c8P2 1 P3 9 5 0
of P4K5 by augmenting A; and (2) improve the global opti-
mality of cost values by evaluating subgroup relationships. We define A = A ∪ 88P1 1 P2 1 P3 99 and the corresponding opti-
Item (1) occurs automatically within the formulation by mal cost (derived by calculating T ∗A in the selection step):
the inclusion of constraints that require at least one new
c ∗A1 n =
X
group not already in A to be generated, ensuring that at c Ai 0
Ai ∈T ∗A1 n
least one next-best product group is appended to A for sub-
sequent steps. Item (2) requires more explanation, which we We could augment A with all members of P4A5 because
illustrate in the following example. the lower-bound estimate can always be calculated from
Splitting a group A into a partition TA allows each sub- singletons. For large problems, P4A5 is prohibitively large;
group of the partition to have its own intermediate batch therefore, we instead augment based on building supersets
(and therefore more flexibility in the choice of mixture com- of existing groups.
position). Global optimality thus requires that We have been careful to call this an estimate of the lower
X bound because nonconvexity could prevent us from accu-
cA ≥ cAi 0 (32) rately calculating the costs of each Ai , because cAi is only an
Ai ∈TA
upper bound. Therefore, a true lower bound cannot be guar-
When we restrict solution sA to the products in Ai ⊆ A (by anteed (although it continues to improve as the algorithm
eliminating any k ∈ A\Ai ), it becomes feasible for any such iterates). Given some desired optimality gap „n , we use
subgroup. It is therefore advantageous to replace sAi by sA „˜ n 2= 4cA1
∗ ∗ ∗
n − cA1 n 5/cA1 n ≤ „n (34)
whenever the condition cA < cAi is detected for any subset
of A. This check is performed for all groups in the solution as a heuristic stopping criterion, with the expectation that
pool A and is motivated by our knowledge that a known the algorithm might terminate prior to achieving true global
locally optimal solution of Ai might be inferior (more costly) optimality within „n .
Esquejo et al.: Laundry Portfolio Optimization at P&G
460 Interfaces 45(5), pp. 444–461, © 2015 INFORMS

Algorithm Summary in 1996 and is a section head in R&D at Procter & Gamble.
Here, we summarize the steps in the algorithm. She has extensive experience in both process and formu-
1. Singleton step. Let iteration count m = 0. Initialize A lation design in the fabric and home care business and is
with all singletons, solve singleton NLP configuration prob- working in the modelling and simulation group, focused on
lems, and calculate similarities ãkl for all pairs of singletons. integration and optimization of models, and application of
2. Grouping step. Let m = m + 1. Solve the grouping big data techniques. She also conducts training and consults
problem and append candidate groups to A, including par- on design of experiments and process control techniques
titions of each group. Also, append to A additional groups internally in P&G.
derived from combinations of members of A. Kevin Miller received his bachelor’s degree in chemistry
3. Configuration step. In parallel, solve independent con- from Xavier University in 1999. He is a principal researcher
figuration problem NLPs for new groups of A. at Procter & Gamble working in Fabric and home care mod-
4. Selection step. Solve set-covering problems for upper elling and simulation. He started in laundry product design
∗ ∗
bounds cA1n and lower-bound estimates cA1n , respectively. for North American Granules products and continued prod-
5. Terminate the algorithm if Inequality (34) holds; else uct design in Central Eastern European and Latin American
go to step 2. Granules. His current focus in modelling and simulation is
on model integration and product optimization.
Kevin Norwood received his PhD in physical chemistry
References from Iowa State University in 1990 and is a research fellow
ASTM International (2014) Standard Guide for Evaluating Stain in R&D at Procter & Gamble. He leads technical work to
Removal Performance in Home Laundering (ASTM International, create and apply modeling approaches to formulate prod-
West Conshohocken, PA). ucts within the fabric and home care businesses. His current
Audet C, Brimberg J, Hansen P, Le Digabel S, Mladenovic N (2004) work is focused on integration of models across disciplines.
Interfaces 2015.45:444-461.

Pooling problem: Alternate formulations and solution meth-


He started with P&G in 1991 and has worked in analytical
ods. Management Sci. 50(6):761–776.
Becerra C (2014) Observed benefits of the portfolio optimization science, technology, formulation, and modeling, where he
approach provided via email communication with Natalie has spent the majority of his career.
Esquejo, June. Ivan Oliveira manages the Advanced Analytics and Op-
Bodington CE, Baker TE (1990) A history of mathematical program- timization Services (AAOS) group at SAS, where he has
ming in the petroleum industry. Interfaces 20(4):117–127. directed projects in operations research (OR) and optimiza-
Box GEP, Stuart Hunter J, Hunter WG (2005) Statistics for
Experimenters2 Design, Innovation, and Discovery, 2nd ed. (John tion applications in a variety of industries. AAOS projects
Wiley & Sons, Hoboken, NJ). deliver consulting expertise to SAS customers in the field
Burnham KP, Anderson DR (2002) Model Selection and Multimodel of OR, inventory optimization, revenue management and
Inference2 A Practical Information-Theoretic Approach, 2nd ed. price optimization, and related technologies. A sample of
(Springer, New York). AAOS projects includes optimal scheduling for ATM cash
Burnham KP, Anderson DR (2004) Multimodel inference: Under-
replenishment, portfolio optimization in government, rev-
standing AIC and BIC in model selection. Sociol. Methods Res.
33(2):261–304. enue management in various industries, retail inventory
Dey S, Gupte A (2013) Analysis of MILP techniques for the pool- replenishment and pricing, chemical mixture portfolio opti-
ing problem. Accessed April 1, 2015, http://www.optimization mization in CPG, optimal assignment of delinquent loan
-online.org/DB_FILE/2013/04/3849.pdf. processing, and simulation for drug discovery. AAOS is
Floudas CA, Aggarwal A (1990) A decomposition strategy for also engaged in internal SAS R&D projects, including opti-
global optimum search in the pooling problem. ORSA J. Com-
put. 2(3):225–235.
mization for data mining. He earned his BS in mechani-
Goos P, Jones B (2011) Optimal Design of Experiments2 A Case Study cal engineering at the University of Virginia and MS and
Approach (John Wiley & Sons, Hoboken, NJ). PhD in mechanical engineering at Massachusetts Institute
Gupte A, Ahmed S, Dey S, Cheon M (2013) Pooling problem: Relax- of Technology.
ations and discretizations. Accessed April 1, 2015, http://www Rob Pratt has worked at SAS since 2000 and is a senior
.optimization-online.org/DB_FILE/2012/10/3658.pdf.
Johnson RT, Montgomery DC, Jones BA (2011) An expository paper
manager in the Operations Research Department within
on optimal design. Quality Engrg. 23(3):287–301. SAS R&D’s Advanced Analytics Division. He manages a
Kutner M, Nachtsheim C, Neter J, Li W (2004) Applied Linear Sta- team of developers responsible for the optimization model-
tistical Models, 5th ed. (McGraw-Hill/Irwin, New York). ing language, network algorithms, and the decomposition
Miller AJ (1990) Subset Selection in Regression (Chapman & Hall, algorithm. He earned a BS in mathematics (with a second
London).
major in English) from the University of Dayton, and both
Visweswaran V, Floudas CA (1990) A global optimization algo-
rithm (GOP) for certain classes of nonconvex NLPs—II. Appli- an MS in mathematics and a PhD in operations research
cations of theory and test problems. Comput. Chemical Engrg. from the University of North Carolina at Chapel Hill.
14(12):1419–1434. Ming Zhao is assistant professor in the Department
of Decision and Information Sciences at the University of
Nats Esquejo received her Bachelor of Science degree in Houston. He served as senior operations research spe-
chemical engineering from the University of the Philippines cialist in Advanced Analytics and Optimization Services
Esquejo et al.: Laundry Portfolio Optimization at P&G
Interfaces 45(5), pp. 444–461, © 2015 INFORMS 461

(AAOS) group at SAS. The projects he has worked on earned his PhD from University at Buffalo in 2008 and
include chemical mixture portfolio optimization for P&G, was a postdoctoral researcher at IBM T.J. Watson Research
operating rooms scheduling, renewable energy integration Center, where he worked primarily on unit commitment
and power system operations, retail inventory replenish- problem and supply chain management in the mining
ment and pricing, and optimization for data mining. He industry.
Interfaces 2015.45:444-461.

You might also like