The Sustainable Development Goals of the United Nations strive to maximize development needs, while minimizing environmental deterioration, without jeopardizing the needs of future generations. Nevertheless, due to urbanization, the escalating trend in natural-resource use, particularly electricity and water, is currently a crucial challenge for sustainable development. One of the promising options is the smart home, which is an extension of building automation with smart characteristics in monitoring, analyzing, controlling, and cloud computing with networked smart devices. Due to the lack of appropriate infrastructure and conscious consumption, its global adoption in the construction industry remains low. We present a technical feasibility of a multi-functional experimental smart home to support the Sustainable Development Goals of the United Nations in terms of water and energy conservation. The layered architecture of the cloud platform with an application program interface enables seamless integration of heterogeneous smart-home technologies and data sources. Use cases demonstrated its capacity to conserve electrical energy and water resources in support of the United Nations’ Sustainable Development Goals. Aside from that, the smart home’s electricity self-consumption of at least three autonomy days was confirmed with zero emissions and electricity bills, and a reduced supply-water consumption.

Badminton is a type of racquet sport that is played using racquets to strike a shuttlecock within a netted rectangular court, and it is one of the most popular sport in Malaysia. In this paper, the team expanded their previous adaptive novel lossless compact view invariant compression technique, named adaptive range of movement index (RoMI) by embedding an intelligent mapping algorithm. Badminton players are broadly classified into two handedness groups, which are left-handed players and right-handed players. The adaptive RoMI technique in the previous module is capable of identifying the labels of normalized RoMI and perform adaptive mapping for badminton players of different handedness. However, the mechanism requires manual judgement from humans to select the type of handedness. This technique will increase the error rate and effort of importing data properties with incorrect manual handedness input or data stream from different handedness datasets of badminton players. Thus, the proposed intelligent adaptive RoMI technique enables the analyzing system to outperform the previous module by detecting the labels of normalized RoMI automatically and by performing intelligent mapping. The motion analysis system will then become much more efficient, intelligent hence simplifying data collection and the handedness invariant benchmarking approach.

Objective: The objective of this study is to investigate Kinect-based badminton movement analysis, with the aim of proposing a lossless compression technique combined with a time-series algorithm system to address the cumbersome and inefficient problems of current analysis methods. Theoretical Framework: This research is underpinned by lossless compression techniques, time-series algorithms (FastDTW), and Kinect-based analysis, which provide a solid basis for the investigation. Method: A system for movement analysis was proposed using a Kinect sensor to capture a dataset of 800 badminton movements from 40 diverse left and right-handed players. An open-source 3D framework was also employed for visualisation. Results and Discussion: The results revealed 96.13% badminton movement recognition accuracy. The framework allows coaches and players to analyse and benchmark movements invariantly with a shorter processing time, regardless of handedness. Research Implications: This research provides data-driven insights to optimise training programmes and enhance athlete performance in the field of sports analysis, using 3D scenes for better movement comparison. It supports SDG 3 (Good Health and Well-being) and SDG 9 (Industry, Innovation, and Infrastructure). Originality/Value: This study addresses a dearth of research in badminton movement analysis by proposing an innovative system to replace inefficient marker-based and video methods. Its value is evidenced by high accuracy (96.13%) and fast, invariant analysis.

Assessing the emotional states of children with learning difficulties (LD) is critical for effective pedagogical and clinical intervention. However, existing diagnostic methods rely on subjective observations which are prone to human bias and lack a localised Asian-centric perspective. This concept paper proposes an AI-based Facial Emotional Recognition (FER) initiative designed to provide objective, real-time assessment of emotional engagement in children with dyslexia during learning activities. The framework utilises a mixed-methods approach, combining a Convolutional Neural Network (CNN) pipeline for video-to-image emotion classification with clinical field validation at the SPARK Child Development Centre. The system leverages 3D depth-sensing technology via Intel RealSense to extract 468 facial landmarks. Preliminary validation of the initiative’s expert-rated dataset achieved a Cohen’s Kappa of 0.87, indicating excellent inter-rater reliability. By bridging the gap between traditional psychiatric screening and modern affective computing, this digitalisation initiative supports the SDG 4 goals, offering a scalable tool for more precise and timely support for children with learning disabilities.

This feasibility study focused on developing a robust Facial Emotion Recognition (FER) system specifically tailored for children aged 7 to 12, aiming to address the limited availability of specialized tools in this field. The system development followed the standard FER pipeline, including face detection, feature extraction, emotion classification, and evaluation. To determine the most effective classification method, three deep learning models, DenseNet-201, ResNet-101, and Inception-v3, were trained using MATLAB. The results showed that DenseNet-201 yielded the highest overall accuracy at 63.10% and was subsequently chosen for integration into the mobile system. The developed system was able to analyze facial expressions frame by frame, offering accurate insights into the emotional states of children. It also supports a user-initiated video processing workflow through an intuitive user interface (UI), minimizing user effort while strictly adhering to privacy protocols.