Leveraging Biometrics for Access Control Management in Learning Institutions

Abstract

A common problem plaguing institutions of higher learning is how to monitor and manage student and staff attendance and access to key location, especially in institutions with multiple locations and the high population. Currently, institutions of higher learning in Kenya use paper-based attendance records coupled with verification of personal documents to control access and monitor attendance, however, this is subject to falsification and inaccurate data. Analysis of such data requires the transfer of the data to tools such as spreadsheets, a process that is subject to errors, missing file, tedious, and time-consuming. This research work focused on coming up with a low-cost scalable biometric identification solution utilizing distributed technology. This would greatly benefit institutions of higher learning seeking to gain accurate and quick facility access and attendance insights on their employees and students. The prototype developed was then evaluated in terms of accuracy and performance to determine its suitability in the identification and processing of data. To test accuracy, 20 participants were enrolled with the system able to effectively verify 97% of the registered users. A simulation of 200 users using JMeter sending HTTP requests through the middleware to the database showed that the system could process data at the rate of up to 450MB per second with a maximum latency of 0.1 seconds. The results demonstrate that the adopted distributed architecture leads to not only low-cost but also an effective system for recording and processing facility access and attendance data.