CENTRALIZED LABORATORY BROWSER ACTIVITY MONITORING SYSTEM

Abstract

Centralized supervision of laboratory computers is difficult when an instructor must inspect each workstation physically. This paper presents a Centralized Laboratory Browser Activity Monitoring System that observes browserwindow titles on distributed laboratory computers and reports flagged activity to a single administrative dashboard. A lightweight Python agent scans visible titles every 15 s, identifies configured applications such as YouTube and ChatGPT, and sends JSON heartbeat messages through a local-area network to a Flask REST backend. The backend updates machine state, creates high-severity alerts, maintains durations, and marks a client offline when no heartbeat is received for 60 s. A React.js dashboard provides live statistics, multi-system status, alert history, per-machine drill-down, charts, agent download, and responsive mobile access. Evaluation on five laboratory workstations reported 12-15 s alert latency, backend response below 50 ms, 1.2 s dashboard loading, agent use below 1% CPU and 28 MB memory, and 100% alert accuracy for the configured activities. The system therefore provides a low-cost, explainable, and minimally intrusive approach for institutional laboratory supervision without requiring commercial endpoint software or cloud infrastructure. Keywords— browser activity monitoring, computer laboratory, Flask REST API, React dashboard, Python monitoring agent, real-time alerts, offline detection.