Volume 25 Issue 12 2025

Responsible farmers are constantly faced with significant financial losses through wild animals that are able to easily get introduced into their agricultural fields which can usually be found close to forests and urban landscapes. We have designed an intelligent Sensor-driven Animal Movement Detection and Adaptive Sound Repellent System (STD-MS) to address this concern. The STD-MS utilizes microwave radar sensors that will detect movement in conjunction to a microcontroller which outputs sounds of different frequencies through speakers. The sounds chosen for the animal will be pitch and loudness adjusted specific to the size and/or behavior of the animal, thus mitigating the chance of habituation. The purpose of this project is to develop an IoT based system for animal detection as well as adaptive sound repellent system for environmentally sustainable crop protection. The system will use microwave radar sensors and microcontrollers to detect animal movement in near real-time within the agricultural fields with wireless IoT monitoring and data logging benefits.

Fingerprint-based biometric systems are widely adopted for secure identity verification. However, they are highly vulnerable to spoofing attacks using fake fingerprints crafted from various materials. To address this challenge, we propose a lightweight feature fusion framework that integrates gradient-based edge information and pore-level micro textures to improve spoof detection accuracy and generalisation. The proposed model uses a convolutional backbone to extract modality-specific features, which are fused via a weighted concatenation mechanism before classification. Experimental results on standard datasets such as LivDet and MSU-FPAD demonstrate that the proposed method achieves superior accuracy (98.1%) compared to state-of-the-art baseline models. Moreover, the system is computationally efficient and interpretable via Grad-CAM-based visualisation, making it suitable for deployment on real-time edge devices. This work complements our prior research by offering a robust yet lightweight defence against spoofing, enhancing the overall integrity of fingerprint biometric systems.

Fused Deposition Modelling (FDM) has emerged as an effective additive manufacturing technique for fabricating polymer and polymer-matrix composites reinforced with various materials. In this study, chopped short carbon fibers were incorporated into polylactic acid pro (PLA PRO) to enhance its mechanical performance, targeting a 30% improvement in overall strength. The composite specimens were fabricated using FDM and evaluated for their tensile strength, flexural strength, and Izod impact strength. Taguchi’s optimization method was employed to determine the optimal combination of printing parameters that maximize these mechanical properties. The influence of key process parameters layer height, printing orientation, extrusion temperature, air gap, and nozzle material was systematically investigated. Results revealed that the careful optimization of these parameters led to substantial improvements in performance. The addition of short carbon fibers (8μm in length and 300μm in diameter) significantly enhanced the tensile strength (61 MPa), flexural strength (70 MPa), and Izod impact strength (480 J/m) of the composite, corresponding to an approximate 30-40% increase compared with pure PLA. Among the investigated parameters, layer height and nozzle material were identified as the most influential factors affecting the mechanical behavior of carbon fiber-reinforced PLA composites.