Integrating Analytics Through the Big Data Information Chain: A Case From Supply Chain Management

• Author: Michael J. Magazine,James W. Hamister,George G. Polak
• DOI: http://doi.org/10.1111/jbl.12192
• Date: 01 September 2018
• Published date: 01 September 2018

Integrating Analytics Through the Big Data Information Chain:

A Case From Supply Chain Management

James W. Hamister

1

, Michael J. Magazine

2

, and George G. Polak

1

1

Wright State University

2

University of Cincinnati

The benefits of good decision making by a distributor often have broad leverage across a supply chain, and data science provides a compre-

hensive framework for making this possible. We present a case study of an ongoing partnership between the authors and corporate man-

agers at a distributor of heating, ventilating, and air-conditioning products. We describe in detail the “vertical integration”of our analytical tools

through a long chain of data scientific activities, backward to raw data, and forward to visually appealing output, in an organization with legacy

information technology infrastructure. The models are applied to a large-scale data set, and spreadsheet-based decision support tools that include

useful visualization capabilities for the firm are illustrated. We also offer this case as a blueprint for building a collaborative research relation-

ship between academia and industry.

Keywords: data science; decision support; inventory management; global optimization; academic–industry collaboration

INTRODUCTION

A distributor often takes a leading decision-making role in a supply

chain and for this reason can wield substantial leverage for reduc-

ing systemwide costs. Achieving this objective, however, requires

the successful completion of a chain of challenging activities that

include data acquisition and cleaning, abstract modeling, populat-

ing the models with realistic data extracted from large files, coding,

and manager-friendly implementation in the field. For the

researcher, the emergent discipline of data science provides a com-

prehensive framework for these activities in much the same way

supply chain management provides a comprehensive approach to

coordinating the productive activities and links in a supply chain.

We present here a case study of an ongoing partnership between

the authors, who are faculty members in academic institutions, and

corporate managers at 2J Supply Company, Inc. The purpose of

the partnership is to apply data science to improve decision making

in sourcing, distribution, and inventory management. The tools

employed range across the areas of predictive, prescriptive, and

descriptive analytics.

We describe in detail the “vertical integration”of our analyti-

cal tools through the long information production chain, as it

were—backward to raw data and forward to visually appealing

output. Along with applications of standard techniques of deci-

sion analysis, our case includes an original model for optimizing

decisions for a distributor that balances mathematical tractability

with the scope of optimization and the formulation of a comple-

mentary algorithmic strategy with appropriate decision support.

Spreadsheet-based mapping tools were developed as well to

enhance the practicability of this approach and clarity of its

results. We also discuss the evolution of our partnership to

include further efforts in predictive and descriptive analytics and

our approach to establishing a long-term relationship with our

industry partners. We also offer this case as a blueprint to read-

ers for building a collaborative research relationship between

academia and industry.

Our case contributes to the literature in data science by detail-

ing an approach to dealing with practical issues that arise in

accessing data and applying optimization and other analytical

tools in an organization with legacy information technology

infrastructure and to making these tools accessible and appealing

to managers. We also offer a new contribution to the stream of

literature on the joint optimization of cycle inventory and net-

work flows by allowing multiple sources and multiple destina-

tions and by formulating inbound and outbound flows in terms

of continuous variables that permit multiple pairings, rather than

in terms of binary assignment variables. Big data analytics can

help a firm realize the potential of optimization by generating the

“knowledge and intelligence to support decision making and

strategic objectives”according to Goes (2014, vi), who further

notes its transformative role in supply chain design. Finally, our

experiences in partnering with 2J Supply have helped us estab-

lish a pattern for forming potential partnerships between the

authors and other businesses, including Crown Equipment,

Honda North America, and Morris Furniture, as well as for other

academics and business managers. In doing so, we contribute to

the literature on academic–industry collaboration (AIC), which is

discussed further in section Academic–industry collaboration.

PARTNERING WITH 2J SUPPLY

The authors are engaged in an ongoing research collaboration

with several managers at 2J Supply, a wholesale distributor of

heating, ventilating, and air-conditioning (HVAC) products in the

Midwest and Greater Ohio Valley in the United States. Prior to

describing the partnership in detail, we review the literature on

AIC.

Corresponding author:

James W. Hamister, Department of Information Systems and Supply

Chain Management, Raj Soin College of Business, Wright State

University, 3640 Colonel Glenn Hwy., Dayton, OH 45435-0001,

USA; E-mail: James.Hamister@Wright.edu

Journal of Business Logistics, 2018, 39(3): 220–230 doi: 10.1111/jbl.12192

© 2018 Council of Supply Chain Management Professionals