Page 155 - UKZN Proceedings of the Conference Report

Testing of the prototype
A series of controlled experiments validated the device’s effectiveness in detecting both steel member theft and electrode degradation or theft.
Steel member monitoring:
• The vibration and audio sensors were tested in a workshop setting. The vibration sensor successfully detected vibrations above the threshold when the angle grinder was used to cut the steel members. The system was also capable of ignoring smaller vibrations from environmental noise, reducing false alarms.
• The audio sensor accurately identified the frequency ranges produced during the cutting of steel, further confirming the presence of theft activity. A back-up method was employed where continuous vibrations over 20 seconds triggered an alert if the audio sensor was inconsistent.
Electrode monitoring:
• The voltage divider circuit was simulated and tested to monitor the footing electrode’s condition. The circuit successfully detected and differentiated between a normal electrode, a degraded electrode (resistance of 20 Ω), and a stolen electrode. Alerts were sent to the
control room via email within 1 minute of detection, providing rapid response capability.
The prototype was able to detect theft attempts and report to control rooms within 1 minute of fault detection, proving the effectiveness of the device in real-time monitoring. The device met the design criteria of low-cost production, long- term sustainability, and minimal false alarms. The total cost for one unit was kept within a budget of R1,600, making it a viable solution for large-scale deployment. This is with consideration of the fact that a large-scale deployment would attract bulk discounts and that the device would be smaller. This will, however, also have to be offset against the cost of a lithium ion battery, which is more expensive than an equivalent lead acid battery. The costing and business case for this specific device is, however, not in the scope of this paper. A successful result may be observed in Figures 6 and 7. A local indication was included on the device (blue LED) to indicate that an event had occurred. This would be useful to field staff when undertaking inspections or confirming the affected tower. In this experiment and angle grinder was used to cut into a steel member that the device was attached to. The blue LED was lit, confirning that an event had occurred, and the email sent (Figure 7) clearly confirms the time of the event at 14:08.
Figure 6: LED illuminating to confirm theft detection and time verification of possible steel member theft incident
Figure 7: Email confirmation of steel member theft being reported to ‘utility control room’.
Proceedings of the conference on Public innovation, develoPment and sustainability | 153