Digitalising Social Housing via Cyber-Physical Systems and AI for Comfort, Health, and Transition to Net-Zero: A Holistic Overview

Decarbonizing the residential sector is challenging due to the complexities in balancing energy consumption optimisation, occupant comfort, and health. Digitalization has emerged as a critical enabler to address these challenges, which could offer data-driven insights to realize this balance and achieve scalable impact, which many existing studies lack. This paper introduces an integrated Cyber-Physical and AI-driven methodology to digitalize Scottish Social housing and tackle these challenges. Our approach employs a low-power Long Range Wide Area Network (LoRaWAN) based internet of Things (IoT) architecture, combining smart plugs, clamp sensors, and environmental sensors to monitor CO₂, temperature, humidity, energy usage at the appliance level, and other critical parameters. Predictive analytics on air quality and energy patterns are also enabled, delivering actionable insights for retrofitting and behavioral optimization. A case study conducted in a Scottish house demonstrated the ability of the proposed system to identify energy inefficiencies. Results show that the system effectively identified energy waste: an average of approximately 30-60% of energy being consumed during unoccupied periods, across multiple appliances. Furthermore, the system achieved over 95% accuracy in predicting CO₂ indoor ppm, enabling relevant proactive actions to future unhealthy air quality conditions. Furthermore, data analytics on thermal fluctuations enabled the identification of poorly insulated rooms, providing insights for actionable retrofitting strategies. The findings highlight the system's ability to reduce energy waste, lower operational costs, and optimize retrofit investments, offering a scalable pathway to achieve low-carbon living standards.