Openserve is using artificial intelligence to separate routine vibrations from unusual activity near its fibre cables and send locations to field teams. The network operator described how physical monitoring fits into a centre overseeing terrestrial, submarine and satellite links. Past simultaneous cable disruptions also prompted transport-capacity upgrades, while old copper infrastructure and equipment compatibility remain constraints on greater automation.
Openserve, the South African wholesale fibre subsidiary of Telkom, is using artificial intelligence and machine learning to flag unusual physical activity near its cables. In the application described to ITWeb, ordinary vibrations from vehicles crossing a bridge can be distinguished from a spade entering the ground nearby. A suspected disturbance receives a location so a human field team can investigate it.[1]
The monitoring centre oversees millions of network elements across terrestrial fibre, submarine connections and geostationary satellite links. It considers optical-network signals, internet-protocol infrastructure, exchanges and environmental temperature. More than 3,000 technicians and subcontractors support operations in the field, and the fibre network extends over 90,000 kilometres. Normal operating patterns provide a baseline for identifying unusual activity.[1]
Sunil Piyarlall, Openserve’s executive for network architecture and modelling, described the system at Conext, a Digital Council Africa industry event. He also recalled submarine disruptions off West Africa alongside cuts in South Africa’s Western Cape and Free State provinces. Losing several paths increased pressure on surviving connections; Openserve subsequently upgraded transport capacity. Automatic rerouting forms another part of operations. Retiring copper infrastructure and connecting equipment from different suppliers remain obstacles to more autonomous management.[1]
