A Review of Artificial Intelligence Techniques for Sustainable Energy Management in Smart Cities: Trends, Challenges, and Future Directions
Year: 2026
Type: article
Abstract: Due to rapid urbanisation, increased energy demands and climate change, cities have evolved to be much more intelligent in managing their energy. This is where Artificial Intelligence (AI) comes into play with sustainable smart cities. AI can be used for many purposes including renewable energy forecasting, smart grid optimisation, demand side management and fault detection. This paper provides a structured review of the literature regarding AI and Machine Learning (ML) used in sustainable energy management within Smart Cities. Twenty-two sources (mainly from 2019 to 2026) were reviewed and split into five categories of Renewable Energy Forecasting, Smart Grid Optimisation, Smart Building and Demand Side Management, Explainable AI and Smart City Sustainability. From this literature, it can be seen that deep learning techniques such as Long Short Term Memory (LSTM) networks, Hybrid Convolutional Neural Network and Long Short Term Memory (CNN-LSTM) and Transformers are dominating in forecasting applications. Reinforcement Learning and Multi-Agent systems are also becoming more popular in real time grid control. Overall though, the literature found that there are many issues surrounding the research and this includes lack of interpretability, scalability, poor quality data and high environmental costs when training large AI models. This paper will outline research gaps, suggest a new conceptual framework around accuracy, interpretability and deployability and suggest future research areas for creating trustworthy, scalable and sustainable AI smart energy systems for Smart Cities.
Authors: Arya Rai, Sujal Paradkar, Mr. Rahul Chaudhari
Language: English
Topic: Smart Grid Energy Management
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Keywords: artificial intelligence, sustainable energy management, smart cities, renewable energy forecasting, demand-side management, fault detection, smart buildings, explainable AI, interpretability, deep learning, LSTM, CNN-LSTM, transformer models, reinforcement learning, multi-agent systems, scalability, trustworthy AI
Open Access status: green
OpenAlex ID: W7236780650
