The rapid improvement of artificial intelligence (AI) in the educational domain has opened new possibilities for enhancing the learning experiences for students. This research discusses the critical need for personalized education in higher education by integrating deep learning (DL) techniques to create customized learning pathways for students. This research intends to bridge the gap between constant educational content and dynamic student needs. This research presents an AI-driven adaptive learning platform implemented across four different courses and 300 students at a university in Faisalabad-Pakistan. A controlled experiment compares student outcomes between those using the AI platform and those undergoing traditional instruction. Quantitative results demonstrate a 25 % improvement in grades, test scores, and engagement for the AI group, with a statistical significance of a p-value of 0.00045. Qualitative feedback highlights enhanced experiences attributed to personalized pathways. The DL analysis of student performance data highlights key parameters, including enhanced learning outcomes and engagement metrices over time. Surveys reveal increased satisfaction compared to one-size-fits-all content. Unlike prior AI research lacking rigorous validation, our methodology and significant results deliver a concrete framework for institutions to implement personalized, AI-driven education at scale. This data-driven approach builds on previous attempts by tying adaptations to actual student needs, yielding measurable improvements in key outcomes. Overall, this work empirically validates that AI platforms leveraging robust analytics to provide customized and adaptive learning can significantly enhance student academic performance, engagement, and satisfaction compared to traditional approaches. These findings have insightful consequences for the future of higher education. The research contributes to the growing demand for AI in education research and provides a practical framework for institutions seeking to implement more adaptive and student-centric teaching methodologies.

Clinical treatment of skin lesion is primarily dependent on timely detection and delimitation of lesion boundaries for accurate cancerous region localization. Prevalence of skin cancer is on the higher side, especially that of melanoma, which is aggressive in nature due to its high metastasis rate. Therefore, timely diagnosis is critical for its treatment before the onset of malignancy. To address this problem, medical imaging is used for the analysis and segmentation of lesion boundaries from dermoscopic images. Various methods have been used, ranging from visual inspection to the textural analysis of the images. However, accuracy of these methods is low for proper clinical treatment because of the sensitivity involved in surgical procedures or drug application. This presents an opportunity to develop an automated model with good accuracy so that it may be used in a clinical setting. This paper proposes an automated method for segmenting lesion boundaries that combines two architectures, the U-Net and the ResNet, collectively called Res-Unet. Moreover, we also used image inpainting for hair removal, which improved the segmentation results significantly. We trained our model on the ISIC 2017 dataset and validated it on the ISIC 2017 test set as well as the PH² dataset. Our proposed model attained a Jaccard Index of 0.772 on the ISIC 2017 test set and 0.854 on the PH² dataset, which are comparable results to the current available state-of-the-art techniques.

Waste management poses a major challenge for cities worldwide, with significant environmental, economic, and social impacts. This paper proposes a novel waste management system leveraging recent advances in the Internet of Things (IoT), algorithms, and cloud analytics to enable more efficient, sustainable, and eco-friendly waste collection and processing in smart cities. An ultrasonic sensor prototype is tailored for reliable fill-level monitoring. A LoRaWAN and cellular network architecture provides city-wide connectivity. A cloud platform handles sensor data storage, processing, and analytics. Dynamic route optimization algorithms minimize time, distance, and fuel use based on real-time bin data. Extensive pilot studies in 10 different locations across Lahore, Pakistan, validated the system, processing over 200 million data points. The results showed a 32% improvement in route efficiency, a 29% decrease in fuel consumption and emissions, a 33% increase in waste processing throughput, and 18% vehicle maintenance savings versus conventional practices. This demonstrates quantifiable benefits across operational, economic, and sustainability dimensions. The proposed IoT-enabled waste management system represents a significant advancement towards sustainable and ecologically responsible waste practices in smart cities worldwide. This research provides a replicable model for holistic smart city solutions integrating sensing, algorithms, and analytics to transition civic operations towards data-driven, efficient paradigms. It represents a significant advancement in sustainable waste practices for smart cities worldwide. Further work could apply emerging technologies like automation and artificial intelligence to create waste management 3.0.

The phenomenal growth of digital learning platforms has brought new learner engagement and retention challenges to higher education. This study proposes a framework that integrates game mechanics—leveling systems, badges, and timely feedback—with artificial intelligence (AI)-driven personalization to meet the challenges of enhanced adaptability, motivation, and learning outcomes in online environments. Key design elements were identified through literature reviews and consultations with instructional design experts, leading to the development an adaptive learning platform prototype. The prototype underwent an eight-week pilot study with 250 Prince Sattam Bin Abdulaziz University (PSAU) students randomly assigned to a control group (non-adaptive system) or an experimental group (adaptive system). Data sources included pre- and post-tests, platform engagement analytics, and learner perception surveys. The results showed that the adaptive group outperformed the control group in the post-test scores (M = 85.2, SD = 6.4 vs. M = 78.5, SD = 7.2) and motivation levels (M = 4.2, SD = 0.7 vs. M = 3.6, SD = 0.8). Additionally, 82% of the adaptive group achieved mastery-level performance compared to 64% in the control group. These findings demonstrate the potential of integrating game mechanics and AI-driven personalization to transform digital learning, offering a roadmap for scalable, data-driven adaptive platforms. Future research will address long-term retention and diverse subject applications.

High voltage direct current (HVDC) power transmission is becoming increasingly important due to steadily rising need for bulk power delivery and interconnected power transmission and distribution systems. DC grids are vulnerable to dc faults, which lead to a rapid rise in dc fault currents. The dc faults must be cleared within timeframe of milliseconds to avoid collapse of the HVDC system. In the event of primary protection (PP) failure, back-up protection (BP) must be applied to clear the fault. In this paper, a novel algorithm based on a Naïve Bayes classifier is proposed to determine threshold levels and operational time frames for primary and back-up protection in multi-terminal voltage source converter based HVDC (VSCHVDC). Local voltage and currents are measured to detect and identify the kind of fault. A four terminal HVDC transmission system is developed in PSCAD/EMTDC and is subjected to lineline faults at different locations and time, to assess the designed protection schemes. Results show that relaying algorithm effectively detects the fault and expedite the primary protection operation. On malfunctioning of PP, BP is accelerated in a short delay of 0.2ms. Furthermore, the relaying algorithm provides faster protection compared to techniques available in the literature. The resulting reduced fault clearance time truncates the maximum fault current and inevitably, leads to reduced power ratings required for dc grid equipment.

Telepresence robots have become popular during the COVID-19 era due to the quarantine measures and the requirement to interact less with other humans. Telepresence robots are helpful in different scenarios, such as healthcare, academia, or the exploration of certain unreachable territories. IoT provides a sensor-based environment wherein robots acquire more precise information about their surroundings. Remote telepresence robots are enabled with more efficient data from IoT sensors, which helps them to compute the data effectively. While navigating in a distant IoT-enabled healthcare environment, there is a possibility of delayed control signals from a teleoperator. We propose a human cooperative telecontrol robotics system in an IoT-sensed healthcare environment. The deep reinforcement learning (DRL)-based deep deterministic policy gradient (DDPG) offered improved control of the telepresence robot to provide assistance to the teleoperator during the delayed communication control signals. The proposed approach can stabilize the system in aid of the teleoperator by taking the delayed signal term out of the main controlling framework, along with the sensed IOT infrastructure. In a dynamic IoT-enabled healthcare context, our suggested approach to operating the telepresence robot can effectively manage the 30 s delayed signal. Simulations and physical experiments in a real-time healthcare environment with human teleoperators demonstrate the implementation of the proposed method.