Design and configuration of a computerized maintenance management system for improving the availability performance of the national ICT broadband backbone: A case of the DAR ES SALAAM–KIMANZICHANA ROUTE, TANZANIA

Abstract

The Dar es Salaam–Kimanzichana route of Tanzania's National ICT Broadband Backbone (NICTBB) has repeatedly fallen short of the 99.999% availability target, with monthly values as low as 73.29% (December 2023) and 86.454% (March 2026). This study identifies the factors behind the shortfalls, examines how logged outage causes relate to monthly availability, and configures a Computerised Maintenance Management System (CMMS) for the route. A mixed-methods design combined a questionnaire given to all 50-route staff (43 returned, 86.0%), 41 months of Tanzania Telecommunications Corporation Limited (TTCL) records and optical time domain reflectometer (OTDR) tests on 40 of 48 fibre cores. Staff perceptions were ranked with the Relative Importance Index (RII), and fibre cuts ranked highest among the technical factors (RII = 0.916). In a regression on 33 months, logged fibre-cut and power-outage hours explained 96.1% of the variation in availability, at about 0.25 points per hour. A model with these two terms predicted eight unseen months with a mean absolute error of 0.24 points. Because both predictors are downtime hours, this result reflects how availability is recorded as much as it explains it. The poweroutage estimate rests for one month, and the configurationresolution effect, which rests for two months, did not hold up and is not used. A CMMS was configured on the Fiix platform with work-order, scheduling, asset, inventory, notification and reporting functions. Network Management System integration and OTDR fault location are planned, and the system has not yet been tested in live use. We recommend completing the integration, backing up power at the nodes, protecting the route and recording outage causes in more detail. 

Key words: National ICT Broadband Backbone, Computerised Maintenance Management System, Availability Performance, Fibre Cuts, Multiple Linear Regression, TTCL.

Received Date: September 17, 2026

Accepted Date: October 09, 2026

Published Date: November 01, 2026

Available Online at: https://www.ijsrisjournal.com/index.php/ojsfiles/article/view/1044