Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Dissertation_GamzeKilincliTaskiran.pdf (3.31 MB)

ETD Abstract Container

Abstract Header

Mathematical Models and Solution Approach for Staff Scheduling with Cross-Training at Call Centers.

Kilincli Taskiran, Gamze
http://rave.ohiolink.edu/etdc/view?acc_num=wright1441028781

Abstract Details

2015, Doctor of Philosophy (PhD), Wright State University, Engineering PhD.
Call centers face demand that varies throughout the week across multiple service categories and typically employ non-standard workforce schedules to meet this demand. In call centers, cross-training provides a buffer against fluctuation of demand between categories and is widely used. Full cross-training, however, is financially impractical in most cases, which has created a challenging problem in how to optimize a cross-trained workforce, i.e., a) what categories should be cross-trained, b) what portion of the workforce should be cross-trained, and c) how to schedule their weekly assignments. This problem is motivated by the need of a Fortune 50 company’s technology support center to schedule its workforce with multiple service categories. To solve this problem to its fullest extent, a mixed integer programming model that addresses staff assignment composition, shift scheduling, days off assignment, and break assignment across multi-skilled agents is proposed. The model is gigantic in size with thousands of general integer variables and is hard to solve. To improve computational efficiency, a two-phase sequential optimization approach is developed. The first phase is to find the optimal composition of the workforce to decide what categories should be cross-trained and when they should be deployed; the second phase is a staff scheduling model to find the size of the workforce with their skill sets and their shifts and weekly tours. The two-phase approach is an order of magnitude faster than the original model and is able to obtain better solutions orders of magnitude faster. Experimental results with real data from the company clearly demonstrate the significance of cross-training; even partial limited cross-training, where 30% - 40% of the workforce is cross-trained with limited (two out of nine) skills per agent, results in considerable performance improvements. The model, when tested in the strategic analysis of the staff composition, suggested an estimated savings of 4% - 9% on staffing cost with an improved service level. Compared with other flexibility options such as part-time shifts, experiment results seem to suggest that cross-training could be a more effective approach to hedge against demand fluctuations when multiple service categories are involved.
Xinhui Zhang, Ph.D. (Advisor)
Pratik Parikh, Ph.D. (Committee Member)
George Polak, Ph.D. (Committee Member)
Yan Liu, Ph.D. (Committee Member)
Nan Kong, Ph.D. (Committee Member)
176 p.
Industrial Engineering
Staff Scheduling, Cross-Training, Integer Program, Optimization

Recommended Citations

Citations

  • Kilincli Taskiran, G. (2015). Mathematical Models and Solution Approach for Staff Scheduling with Cross-Training at Call Centers. [Doctoral dissertation, Wright State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=wright1441028781

    APA Style (7th edition)

  • Kilincli Taskiran, Gamze. Mathematical Models and Solution Approach for Staff Scheduling with Cross-Training at Call Centers. . 2015. Wright State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=wright1441028781.

    MLA Style (8th edition)

  • Kilincli Taskiran, Gamze. "Mathematical Models and Solution Approach for Staff Scheduling with Cross-Training at Call Centers. ." Doctoral dissertation, Wright State University, 2015. http://rave.ohiolink.edu/etdc/view?acc_num=wright1441028781

    Chicago Manual of Style (17th edition)

wright1441028781
3,263
© 2015, all rights reserved.
This open access ETD is published by Wright State University and OhioLINK.