Applied Mechanics and Materials Vol. 935

Abstract: Nowadays, many people have difficulty accessing dedicated exercise facilities in the gymnasium. This could be due to hectic work schedules or constant travel. In such a scenario, a person will struggle to stick to an exercise routine to keep fit. Though people have adopted electronic devices for fitness tracking, such devices usually rely on batteries and must be worn in most cases to monitor fitness activities. Hence, we proposed the design of Fit-Track, a battery-free and wearable device-free fitness tracking system with edge intelligence. Fit-Track proposes using piezoresistive (PZR) sensors to track specific exercises (such as steps, pushups, and squats). The PZR sensors are interfaced with Arduino nano33 BLE sense for data processing and wireless transmission to a mobile phone. During the design of Fit-Track, we faced challenges of power supply, interference, and computation. Those challenges were addressed through adopting a multi-modal (solar and kinetic energy harvesting) power supply, attaching a PZR sensor to each foot, and developing algorithms (both intelligent and signal processing) based on the nature of the exercise carried out for tracking counts of the activity. Experiments on Fit-Track for the various exercises (steps, push-ups, and squats) with edge intelligence showed a minimum accuracy of 99 %.

Abstract: In recent times, electricity has become a great necessity for everyone and almost everything around the globe. The ever-growing population and increase in electronic devices have increased the rate of energy consumption. In a bid to meet this demand, all forms of energy are utilized and still undergoing research. This work aims at generating and harvesting electrical energy by converting mechanical energy from pressure exerted (footsteps) on the material into electricity by a phenomenon called piezoelectricity. A group of 40 piezoelectric crystals were constructed and connected via a parallel connection of 5 by 8. The piezoelectric crystals showed different results to varying weights and oscillations at a fixed time. These results were visible on the plot of voltage generated versus weight and on the plot of voltage generated to number of steps. These graphs showed that the greater the weight the more voltage and also more voltage is generated with increase in number of footsteps.

Abstract: This study investigates the deployment of adaptive neural network-based control strategies for nonlinear dynamic systems, emphasizing the integration of Echo State Networks (ESNs) into a feedforward-feedback control architecture. Traditional controllers relying on precise mathematical modeling often fail to cope with the complexity of systems exhibiting high nonlinearity, time-varying parameters, and external disturbances. The proposed ESN-based approach harnesses reservoir computing to construct a lightweight, data-driven model capable of accurately capturing system dynamics in real time. The feedforward module provides anticipatory control actions, while the feedback loop compensates for deviations, enabling rapid convergence and robustness against parametric drift. Comparative analysis with conventional PID and LQR controllers reveals superior performance in terms of tracking accuracy, stability, and noise resilience. Preliminary simulations predict reduced steady-state error and improved dynamic response even under uncertain operating conditions. This architecture presents a scalable and efficient alternative for advanced applications in robotics, aerospace, and industrial process control. The findings affirm the viability of ESNs in redefining adaptive control paradigms by combining interpretability, computational efficiency, and real-world adaptability. Reference to this paper should be made as follows:MCE 2025, MCE825. (2025) ‘Adaptive neural network-based feedforward-feedback controller for nonlinear dynamic systems.

Abstract: This paper presents the design, construction and development of an autonomous delivery robot aimed at structured environment such as university campuses, hospitals, residential estates and factories. The system integrates a six wheeled differential drive platform with a redesign adaptive climbing rocker-boogie suspension system. It makes use of an array of high precision sensors such as LIDAR, ultrasonic sensors, IR sensors, depth camera, real time kinematics (RTK) GPS for real time navigation and obstacle detection. The autonomous delivery robot is managed using ROS 2-based system running on an Nvidia Jetson nanoand features a mobile application for remote tracking, management and control. Simulation based testing in gazebo as well as experimental validation was conducted to evaluate the robot’s autonomous behavior and delivery performance.

Abstract: Path planning refers to designing a reliable, feasible, optimum, safe, and collision-free path with the shortest distance that takes a mobile robot from the start position to the goal point within an environment. To ensure the successful operation of a robot, an effective and efficient Path planning technique that guarantees obstacle avoidance and an optimal path must be adopted. This paper applies a novel path length reduction technique – the Kenneth, Nnanna, and Saleh (KNS) algorithm-to notable path planning algorithms (APF, A*, RRT, and RRT*) to shorten their path length by reducing the waypoints' bends and retaining the obstacle avoidance capability of the algorithms. We simulated applying the technique to different notable algorithms in an environment configured with varying obstacles. We compared the resultant paths with the original paths. The results show that the KNS algorithm is very effective and can significantly reduce the path length of the notable algorithms.

Abstract: Mechanical vibrations are abundant in human-made environments and can be harnessed using piezoelectric transduction. Among the piezo materials, piezoelectric polymers exhibit flexibility and mechanical compliance, improving resilience to shock and deformation—suited for low-frequency high strain environments. In this paper, distinct designs of piezoelectric active area were topology optimized using ANSYS. Three designs of bimorph cantilevered energy harvesters were developed to obtain the optimum material layouts of piezoelectric PVDF, maximize the voltage output, decrease the resonant frequency, and reduce the amount of material needed. Two additional designs with varying volume retainment were also simulated to investigate the effects of optimization parameters. The best topology optimized design, #2, had a resonant frequency of 16.9 Hz and a piezo voltage of 1.08E-3 V/mm³ normalized to the amount of remaining PVDF after optimization. Although the frequency is still higher than the target ambient energy sources, this study showed that topology optimization in conjunction with design can be used to define structures leading to the energy harvesting application frequency.

Abstract: Detecting microscopic defects in precision manufacturing remains a major challenge, particularly in hard disk drive (HDD) production where sub-millimeter dust particles on the Voice Coil Motor Assembly (VCMA) can cause performance degradation or early device failure. This study presents a comparative evaluation of three YOLO object-detection architectures—YOLOv5, YOLOv8, and YOLOv11—applied to high-resolution dust detection on VCMA components. All models were trained and tested using the same annotated 5-megapixel dataset under identical experimental settings to ensure fair comparison. The results show that YOLOv5 achieved the highest precision (0.640) and the highest mAP50–95 (0.253), indicating stable localization performance across strict IoU thresholds. YOLOv8 produced the highest mAP50 (0.500), reflecting strong localization accuracy at IoU 0.5, while maintaining moderate precision (0.633) and lower recall (0.455). YOLOv11 obtained the highest recall (0.636), successfully capturing the largest proportion of true dust particles, though with lower precision (0.335) and weaker mAP values, revealing a higher rate of false detections. Overall, the findings highlight clear trade-offs among the models: YOLOv5 offers the most balanced performance, YOLOv8 excels in spatial localization, and YOLOv11 is suitable for scenarios where maximum defect coverage is prioritized. These insights support the selection of appropriate detection architectures for automated micro-defect inspection and contribute to the development of AI-driven quality-control systems in HDD manufacturing.

Abstract: This paper describes a development of Electrical Discharge Machining (EDM) system for biomedical application. In general, the mechanism of EDM comprises of mechanical structure and electronic control system. This laboratory scale of the EDM system has a capability to accommodate the machining of hip implant which employs low power generator. The holder for the workpiece is created to accurately position the hip implant, ensuring that the machining angle of the implant directs the micro-pits precisely toward the workpiece. A traditional linear x-y-z axis setup (Cartesian coordinate system) is utilized, along with two types of spherical coordinates (swing-swing and swing-rotate configurations). By the results of performance test, the Swing Motor behaves differently to the common servo motor. The Swing Motor is affected by unbalanced load and gravity in which the Ziegler-Nichols PID optimization method has been altered from the conventional model. The average of absolute error is 0.2308 degrees. However, optimized PI controller by Ziegler-Nichols method is able to eliminate the effect in term of final achieved position (steady state condition) and fulfil the objectives.

Abstract: Emerging designs of devices require sophisticated bond pad architecture to meet certain specifications, design applications, as well as package requirements. Sophisticated bond pad structures often have thin metal layers and POA circuit bond pads underneath which require careful application of wire bond processing to avoid cracking on the bonding pads during wire bond. Bond pad crack is one of the most detrimental issues at wire bonding, especially with POA devices, so it is important to take into consideration the wire material to be used, the process parameter to be defined, as well as the structure of the bonding pad. This paper aims to resolve and eliminate the bond pad crack by the application Initial Force via Force Profiling, whilst adhering to the output response criteria at wire bond, and not going outside the defined process parameter window. Furthermore, this paper aims to help readers to have a more comprehensive understanding of the Force parameters at wire bond, as well as the different architecture of bonding pads.

Abstract: 3D printing parameters such as printing temperatures and speeds play a vital role in the melt flow and printability of thermoplastic filaments in fused filament fabrication (FFF) technology. Inappropriate print settings mainly induce incomplete and poor printing quality due to melt flow instability. This research work focused on modeling the melt flow behavior of polylactic acid (PLA) at different printing temperatures and speeds using computer fluid dynamics (CFD) method. The shear stress and viscosity of PLA were investigated by a melt flow indexer (MFI) and rheometer in temperature ranges of 200 - 240 °C. A model of a capillary tube in MFI was set up with an initial condition of rheological properties from the experiment to simulate the hot melt extrusion relating to the melt flowability of PLA filaments. The high shear stress and low viscosity presented at the edge of filaments at every printing condition. Additionally, the shear stress and viscosity decreased linearly when the printing temperature increased, while the shear stress increased when the printing speed increased. The increase in shear stress caused high surface roughness of PLA specimens after printing. The findings can guide the optimization of the FFF 3D printing process to improve surface finish quality.