Owen Smith

Abstract

The methods of operation of many newspaper camera departments have been established over a long period by many different supervisors and department heads. The system may have been good once but went bad because the nature of the department's operations changed without corresponding changes in the system. The production manager must be able to detect when a system needs modifying. Presently, there are few tools available to the manager for evaluating the effectiveness of the department. The goal of this study was to demonstrate how a very simple analytical technique could be applied to a camera department's operations and, thereby, determine the department's operating characteristics. This analytical technique is called queueing analysis. The particular camera department studied contained two independent systems. One camera serviced pasted-up pages while the other serviced halftones and artwork. Each of these systems or channels was studied separately. However, before the systems could be studied, several assumptions of the queueing model had to be satisfied. These assumptions were: (1) copy arrivals must be Poisson distributed, (2) copy service time must be exponentially distributed, (3) copy can not leave the waiting line after having entered, (4) the service discipline must be first-in-first-out, (5) and the arrival pattern must be random. The only raw data required by the analysis were the copy arrival time and the service time of each piece of copy. The data were compiled to form empirical distributions which were compared to theoretical distributions to determine the feasibility of the assumptions. The average arrival rate and service rate were also determined from this compiled data. These two averages were substituted into formulas developed by theorists to determine the operating characteristics or measures of effectiveness for the camera department. The analysis was carried out over a two week period and considered each system on a weekly basis. In this manner the two weeks of operation could be compared. It was found that the page channel could be monitored quite easily using queueing analysis, but the halftone/artwork camera could not be studied without modifications. It was found that the expected time a page spent in the system was much longer than previously thought. During the first week of observation, the expected time in the system was 18.9 minutes per page. It was 11.4 minutes per page during the second week. There was also a great difference between the two times which might indicate that something in the system had changed. It was found that the expected number of pages in the system per half hour during the first week was 1.69 and 1.06 during the second week. It would be expected that the system could handle a greater number of pages than this. This observation was further supported by the weekly busy period probabilities determined for the system. This is the probability that the system will be busy during any half hour interval during the week. The busy period probability was .63 for the first week and .51 for the second. These figures indicated that there was a great deal of idle time in this system. In fact, there was 23 hours and 19 minutes of idle time during the first week and 30 hours and 52 minutes during the second week. It was impossible to draw any straightforward conclusions from these results without knowing the objectives of the organization. If management's goal was to minimize costs in this department, it was evident this objective was not being met in light of the great amount of idle time that existed. To realize this goal, management must reduce this idle time. If management's goal was to maximize speed, the system would again be out of line with management's expectations. However, this goal may have been more closely realized. To fully satisfy this objective would require management to reduce the time copy spends in the system. This study demonstrates that queueing analysis can be used quite successfully and easily in determining the operating characteristics of a camera department. It was also demonstrated that even when the analysis was not successful, much could be learned about the system through the application of this tool.

Library of Congress Subject Headings

Photojournalism; Queuing theory

Publication Date

5-1-1978

Document Type

Thesis

Department, Program, or Center

School of Media Sciences (CIAS)

Advisor

Craig, W. Frederick

Advisor/Committee Member

Hacker, Robert

Comments

Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: TR820.S56

Campus

RIT – Main Campus

Download

Share

COinS