11.08.2013 Views

CALL CENTERS (CENTRES) - Faculty of Industrial Engineering and ...

CALL CENTERS (CENTRES) - Faculty of Industrial Engineering and ...

CALL CENTERS (CENTRES) - Faculty of Industrial Engineering and ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Abstract. Concerns the development for mail order company L.L. Bean <strong>of</strong> a computerized<br />

procedure for selecting complex large-scale telephone-operator scheduling systems. To assess<br />

capability in forecasting work load, setting requisite capacity levels, <strong>and</strong> generating satisfactory<br />

work-shift schedules, cost/benefit analysis was used <strong>and</strong> the expected penalty costs <strong>of</strong> lost orders<br />

due to understaffing <strong>and</strong> loaded-wage costs <strong>of</strong> overstaffing were considered. Queuing theory was<br />

used to model customer-call behavior for every hour over 24-hour days, seven days per week, <strong>and</strong><br />

the results <strong>of</strong> linear regression, which correlated customer-service level with expected customer<br />

ab<strong>and</strong>onment rate, were implemented to estimate the impact on order revenues <strong>of</strong> telephoneservice<br />

level.<br />

Keywords: Work load forecasting, Queueing theory, Telephone agent scheduling system, Mailorder<br />

company, L.L. Bean, Computerized procedure, Complex large-scale telephone operator<br />

scheduling systems, Capacity levels, Work shift schedules, Cost-benefit analysis, Expected<br />

penalty costs, Understaffing, Loaded wage costs, Overstaffing, Linear regression<br />

(Appears also in Section VIII.)<br />

23. Green, L. <strong>and</strong> P. Kolesar. Testing the validity <strong>of</strong> a queueing model <strong>of</strong> police patrol, Management<br />

Science, 35 (2), 1989, 127–148.<br />

Abstract. This paper describes efforts to validate a multiple car dispatch queueing (MCD)<br />

model <strong>of</strong> police patrol operations using New York City data. The MCD model was designed for<br />

use in a computer system that has been disseminated to many police departments in the U.S.<br />

to help planners allocate patrol cars among precincts. It has also been used to evaluate specific<br />

changes in patrol policy in New York. We define validation as a series <strong>of</strong> hierarchical procedures<br />

ranging from tests <strong>of</strong> mathematical correctness to evaluations <strong>of</strong> model robustness. We discuss<br />

the difficulties <strong>and</strong> limitations <strong>of</strong> assessing the validity <strong>of</strong> a model <strong>of</strong> a loosely managed system<br />

in which human behavior is central <strong>and</strong> in which controlled experiments cannot be performed.<br />

Focusing on specific uses, we conclude that the MCD model is a good although imperfect description<br />

<strong>of</strong> patrol operations in New York <strong>and</strong> is a potentially useful planning tool for many<br />

other urban police departments.<br />

Keywords: Queueing, Validation, Police patrol, Multiple car dispatch queueing model, New<br />

York City, Hierarchical procedures, Mathematical correctness, Model robustness, Urban police<br />

departments<br />

24. Blake, R.H., S.C. Graves <strong>and</strong> P.C. Santos. A model for the configuration <strong>of</strong> incoming WATS<br />

lines, QUESTA, 7, 1990, 3–21.<br />

Abstract. WearGuard is a direct marketer <strong>and</strong> retailer <strong>of</strong> uniforms <strong>and</strong> work clothes, which<br />

relies primarily on phone orders for sales. For this purpose it maintains a series <strong>of</strong> toll-free “800number”<br />

lines, known as WATS lines, to receive its incoming calls. These lines are <strong>of</strong> several<br />

types, where each type serves a different portion <strong>of</strong> the country <strong>and</strong> has a different usage fee.<br />

In this paper, we determine how many <strong>of</strong> each type <strong>of</strong> WATS lines should be employed. After<br />

defining the problem more completely, we develop a queueing model to describe the system <strong>and</strong> a<br />

dynamic program to solve the configuration problem to optimality. The model has been applied<br />

to the problem by WearGuard since 1984. We present an example <strong>and</strong> examine the sensitivity<br />

<strong>of</strong> the solution to variations in various parameters. We validate the model by comparing the<br />

9

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!