With up to 65%1 of Southeast Asia’s population — more than 500 million people2 — expected to live in urban areas and electricity demand set to triple3 by 2050, the reliability of power distribution networks is essential to support economic growth in the region. Accelerating urbanization is driving the need for greater reliance on underground power cables, especially where more robust, space-efficient power distribution is needed and overhead lines are impractical.
Increased cable failures during heavy rain or wet seasons are common in Southeast Asia, raising the risk of outages during peak demand periods. While underground power cables offer advantages, such as reduced exposure to weather-related disruptions and aesthetic benefits, they are not immune to failure. Aging infrastructure, insulation degradation, defects due to improper handling and environmental stressors are other factors contributing to unplanned cable outages that disrupt services, increase operational costs and frustrate consumers.
To manage these challenges, predictive analytics, artificial intelligence (AI) and machine learning (ML) could be game-changing for power distribution. Such technologies can be used to help utility companies anticipate underground cable failures before they occur, optimize maintenance schedules and enhance grid resilience.
Use of predictive analytics in underground cable management
Predictive analytics is fundamentally transforming how utilities manage underground cable networks. By leveraging historical data and real-time information collected from the Internet of Things (IoT) sensors embedded in cables, utilities can identify patterns that signal potential failures. This shift from reactive to proactive maintenance has demonstrated measurable success in various parts of the world.
For instance, a utility company in France has been using AI for several years to help optimize network renewal decisions. Initially applied to low-voltage underground networks, their machine learning models are now also used in medium-voltage networks. By analyzing historical fault data and features like cable age and insulation type, the company can successfully identify high-risk feeders for replacement or maintenance. This approach allows it to prioritize investments effectively across its extensive underground cable networks for resilience during extreme weather events like heat waves.
Similarly, a utility company in the US has implemented targeted undergrounding programs using advanced data analytics. These initiatives strategically identify outage-prone overhead lines and convert them into underground systems, resulting in reduced outage durations for residential and commercial customers during hurricanes and other extreme weather events.
Real-world applications of AI and ML in cable maintenance
AI and ML take predictive analytics a step further by automating the analysis of complex data sets and uncovering insights that would be impossible for humans to discern manually. These technologies analyze parameters like temperature fluctuations, vibration patterns, soil conditions and electrical loads to help predict when and where a cable might fail.
A utility company in Australia has a predictive maintenance program incorporating AI tools that simulate cable aging under various environmental conditions. These simulations allow the company to test different scenarios virtually before implementing solutions on the ground — a cost-effective approach that helps reduce trial-and-error efforts.
In the UK, a utility company has gone a step further by integrating predictive analytics into its broader digital grid strategy. The company combines AI insights with self-healing technologies that automatically reroute power during outages, reducing downtime for customers while maintaining service continuity across its network.
Grid reliability and customer satisfaction
By helping to prevent outages, predictive analytics offers significant advantages for residential and commercial customers who rely on an uninterrupted power supply for their daily lives and operations. Moreover, proactive maintenance greatly reduces the need for emergency repairs that often result in prolonged service interruptions. Scheduled maintenance activities are less disruptive and more predictable, allowing customers to plan accordingly. This shift aligns with growing consumer expectations for consistent service quality in an increasingly digital world.
By optimizing maintenance schedules and reducing unplanned outages, utilities can lower their operational expenses. These savings can then be passed on to consumers through stable or even reduced electricity tariffs over time.
Enhancing sustainability efforts
Predictive analytics and AI-driven cable management could contribute significantly to sustainability goals — a priority for utility companies worldwide as they navigate the global energy transition. Preventing underground cable outages reduces reliance on backup generators that often run on fossil fuels like diesel or natural gas during emergencies. By greatly reducing the use of these carbon-intensive solutions, utility companies can cut greenhouse gas emissions associated with outage management.
In addition, predictive maintenance reduces the need for emergency repair operations involving heavy machinery and transportation fleets, which are also sources of emissions. Over time, these reductions can help lower the organization’s overall carbon footprint.
Importantly, reliable grids are essential for integrating renewable energy sources like solar and wind into the energy mix. Intermittent renewables require stable distribution networks capable of handling variable loads without frequent disruptions. By driving grid reliability through predictive analytics and AI, utility companies can accelerate their adoption of clean energy technologies — a critical step toward achieving net-zero emissions targets.
