Page 61 - UKZN Proceedings of the Conference Report

aI-driven innovations for South africa’s multidimensional challenges: energy optimi- sation, crime reduction, economic opportuni- ties, and environmental sustainability
Energy management and optimisation
Considering the historical dependence on coal, the South African energy sector is in an interesting position to embrace AI technology in its quest to solve the longstanding crisis. Inherent load shedding and intermittent electricity supply have crippled economic growth for decades. One of the areas where AI can be used to replace energy analysts is in the case of demand forecasting. For instance, AI can help narrow down and create a more sophisticated pump management system by using information on currents and historical data, temperature trends, and relative humidity levels. A case in point is the activities of a Smart City in Dubai, which is a case study of how AI is used in harnessing renewable energy sources (Go-Globe 2024). The Smart Grid Initiative, indeed, represents one of the chief innovations toward establishing a modern and sustainable energy system in Dubai. It promotes the use of electric vehicles (EVs) on the one hand and increases energy efficiency by integrating renewable sources on the other. Furthermore, sustainability is further supported by the integration of biomass into this process (Davis 2024). This in turn creates significant potential for market players dealing with clean tech and green mobility solutions. Furthermore, smart grids accompanied by modern metering systems and demand side management systems encourage new developments in managed resources and smart home- related services.
The Dubai Electricity and Water Authority (DEWA) was the first to implement intelligent networks using AI algorithms to study their clients, which leads to more optimised power consumption through informed resource distribution. As seen in Mega Trends and Analysis (2022), DEWA’s strategies are a process of changes aimed at turning Dubai into the smartest city in the world. A smart grid is one of the modernisation components, which consists of contemporary features that facilitate operational decisions in real time and also connects seamlessly with the water and electricity infrastructure. DEWA has made significant achievements in delivering on the objectives contained in its Smart Grid Strategy implemented for 2014 to 2035. Between 2015 and 2020 full automation of its deviation network was achieved, which is associated with 400kV and 132kV substations. Over two million two-way interactive smart meters for water and electricity, which help consumers to optimise their energy usage, were deployed. DEWA also expanded its multi-application Radio Frequency (RF) mesh network and established communication with more than 4,200 distribution substations within Dubai (Mega Trends and Analysis 2022).
Likewise, the United Kingdom is no exception in attaining progress in adopting energy- efficiency practices employing app technologies. Predictive modelling using AI and machine learning incorporated by the national grid seeks to optimise the energy outputs of solar and wind gas, especially where changing weather conditions such as fluctuations in wind and sunlight are present. In partnership with the Alan Turing Institute, the National Grid Electricity System Operator trained AI prediction models, which enhanced solar-forecasting performance by one-third (Cuff 2019). Supply and demand in this ‘suitable’ manner has made it possible for the United Kingdom to improve on the efficiency of consumables as well as prices charged to consumers. Such innovation is in line with the UK’s strategy of providing entirely carbon-free generation of electrical energy within the period of the next six years (Cuff 2019).
From a sociological standpoint, it is necessary to observe independence from renewable energy through examples to appreciate the challenges involved in the deployment of renewable energy technologies. For instance, excessive community opposition was recorded for some wind energy projects, causing them to be cancelled, which underlines the role of the communities’ right of acceptance (Petrakopoulou 2017). This is also the case in China and Iran, where retraining is required due to the underdevelopment of technical capabilities such as efficient renewable resource management (Liu et al. 2011; Ghorashi and Rahim 2011). Even so, AI, which enhances operational efficiency of energy-generating systems such as solar panels and wind turbines by enabling predictive maintenance, has been adopted. Devices are structurally equipped with data analytical applications that can scan changes within a period of information collected on temperatures, performances, and weather, well in advance of a breakdown. With this in effect, downtime is limited and energy production is at a peak, hence increasing efficiency in energy systems operations (Evans et al. 2009). In addition, as illustrated in the work of Cuevas and Gonzalez (2013), it is also suggested that the use of AI is also imperative in enhancing electricity generation by enhancing resource management tasks such as cloud cover prediction on solar panels and wind pattern for turbine refurbishing.
The incorporation of renewable energy technologies into pre-existing power systems is the greatest challenge, mainly due to the variable nature of renewable energy resources. Nevertheless, AI proves invaluable in improving grid stability and management. By forecasting energy production, AI allows for better management of the energy supply through enhanced coordination of energy production and demand (Lund 2007). Besides supporting management of the grid, AI also has an effect on the
Proceedings of the conference on Public innovation, develoPment and sustainability | 59