Title:
Service Engineering: Data-Based Science in Support of Service Management, or Empirical Adventures in Call Centers and Hospitals
Service Engineering: Data-Based Science in Support of Service Management, or Empirical Adventures in Call Centers and Hospitals
Author(s)
Mandelbaum, Avishai
Advisor(s)
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
Synopsis: In this lecture, Dr. Mandelbaum will describe examples of
complex service operations for which data-based simple models have been
found useful, which he refers to as "Simple Models at the Service of Complex
Realities." Examples include call centers, hospitals, banks, courts and more.
He views these service systems through the mathematical lenses of Queueing
Science, with a bias towards Statistics.
The mathematical framework for his models is asymptotic queueing theory,
where limits are taken as the number of servers increases indefinitely, in
ways that maintain a sought-after (often delicate) balance between staffing
level and offered-load. A specific such balance reveals an operational regime
that achieves, under already moderate scale, remarkably high levels of both
service quality and efficiency. This is the QED regime (Quality- & Efficiency-
Driven), discovered by Erlang and substantiated mathematically by Halfin &
Whitt.
The data-source for the lecture is a unique data repository from call centers
and hospitals. The data is maintained at the Technion's SEE Laboratory
(Service Enterprise Engineering). It is unique in that it is transaction-based;
it details the individual operational history of all the service transactions
(e.g., calls in a call center or patients in an emergency department). One
source of data, publicly available, is a network of four call centers of a U.S.
bank, spanning two and a half years and covering about 1,000 agents; there
are 218,047,488 telephone calls overall, out of which 41,646,142 were
served by agents, while the rest were handled by answering machines. The
data can be explored via SEEStat, an environment for online Exploratory Data
Analysis. SEEStat is accessible here after registration.
Sponsor
Date Issued
2011-02-24
Extent
70:33 minutes
Resource Type
Moving Image
Resource Subtype
Lecture