Engineering Headway Vol. 28

Abstract: The global wood industry generates substantial volumes of residues that require effective utilization, marketing, and disposal. Over 106 tons of wood residue was exported in 2021 in Namibia. However, comprehensive data on Namibia's wood waste generation, utilization, and disposal of unutilized waste are unavailable. Lack of empirical data on wood waste management specific to Namibia may hinder the development of effective and sustainable management strategies necessary for mitigating environmental impact and protection of public health. This study aims to quantify the volume of wood residue generated; and identify the waste management practices employed by Namibian wood processors, sawmills, and plank vendors. Survey design, data collection through interviews and structured questionnaires in selected locations within Namibia was adopted in this study. Statistical analysis conducted shows that an average of 5.5% of wood waste remains unused amounting to 43.79 tons annually for the selected areas and facilities. The amount of wood waste from the selected facilities was evaluated to be 100.52 m³/month. The results also help to identify significant environmental, and health related problems associated with wood waste management and possible mitigation strategies. Alternative uses for wood waste as a means of promoting a circular economy in Namibia is herein recommended in this study.

Abstract: Pilot injection is a useful technique for lowering high rates of pressure rise at injection initiation and ignition delay. This leads to increased hydrocarbon emissions but less nitrogen oxides and noise. This work considers the impact of pilot injection on a twin-cylinder turbocharged common rail direct injection diesel engine powered by diesel. By adjusting the pilot injection quantity from 5 to 25% (in increments of 5%) of the main injection quantity and the pilot injection timing from 0 to 25 crank angle degrees (CAD) in steps of 5 CAD before the main injection, engine experiments were conducted at a constant speed of 2500 rpm and fuel injection pressure of 120 MPa. Performance factors like brake thermal efficiency and brake specific fuel consumption (BSFC), and emission factors like soot and nitrogen oxides were assessed experimentally. The findings show that while BSFC and soot decreased with a rise in pilot injection quantity, BTE and NOx increased. Prior to the main injection, at a pilot injection timing of 10 CAD, improved BTE and BSFC were noted. In relation to pilot injection timing, NOx and soot emissions exhibit opposing tendencies. While soot emissions initially decline until a pilot injection timing of 15 CAD, and then rise, NOx emissions first rise until a pilot injection timing of 15 CAD and then fall.

Abstract: Manual potato harvesting is labour-intensive and time-consuming for small-scale farmers due to the high cost and inaccessibility of industrial machinery. This study presents the development of a small-scale potato harvester that integrates digging and sieving mechanisms. Three design concepts were assessed based on capital investment, size, maintenance, effectiveness, and digger adjustability. A detailed design analysis determined key specifications, including machine dimensions (2.385 m × 1.07 m × 1.083 m), cutting share width (0.51 m), weight (180 kg), field capacity (0.101 ha/h), and operating speed (3.5 km/h). The digger is designed to rotate for adaptability to varying soil slopes and is vertically adjustable for different depths. The estimated cost of the developed harvester is approximately N$30,000, making it a cost-effective alternative to industrial machines. The designed mechanized potato-harvesting machine is expected to significantly reduce labour, improve overall efficiency, and enhance productivity, thereby easing the harvesting process for small-scale farmers. The isometric view of the potato harvester, designed in SOLIDWORKS®, illustrates the final design.

Abstract: The detonation cell size is an important parameter for evaluating gas detonations and developing techniques for explosion damage mitigation. The resource-intensive experiments on gas detonation are carried out under limited test conditions, often leading to interpolating, or extrapolating from available datasets. Existing detonation cell width prediction models using Machine Learning models either use chemical-based input parameter computed using Cantera or CHEMKIN II packages. Parameters defined by chemical mechanisms introduce both, model errors and parameter uncertainties that could affect the prediction accuracy. In the present paper, experimental gaseous detonation data for hydrogen-oxygen and hydrogen-air mixtures are statistically analyzed to establish features for detonation cell widths prediction using machine learning models. Machine learning models are trained and validated using experimental data available from literature and internal tests. SHapley Additive exPlanations method is used for feature impacts analysis on model predictions. For non-diluted mixtures, detonation cell widths prediction based on initial mixture composition, pressure, and temperature are made for hydrogen-air mixtures (mean absolute error of 0.02 mm) and for stoichiometric hydrogen-oxygen mixtures (mean absolute error of 0.16 mm) at an averaged 99% accuracy. The models’ performance is validated against most recent models and new datasets and the experimentally reported detonation cell width measurements uncertainties.

Abstract: The challenges of vision sensors during poor visibility or when faced with transparent objects, along with RADAR sensors’ susceptibility to degradation and jamming constitute major aspects of limitations in object detection systems. Researchers are continuously exploring better methods and materials for object detection to develop fully autonomous vehicles capable of avoiding collisions in scenarios where distance and speed permit. This paper examines the accuracy of measurements taken by the Level metre Ultrasonic sensor, range finder LiDAR and a radar sensor when targeting objects with contrasting levels of visibility. The target objects considered for this experiment are black and light brown coloured plywood and transparent glass; these objects represent opposites in terms of visibility. Results for range finder LiDAR showed minimal variations with the plywood objects, but recorded much higher error margins when targeting the glass object and failed to detect the glass object at a range of 0.6 metre; which was the shortest distance considered for this paper. The ultrasonic sensor recorded an average error margin of 2.03% at distances between 0.6 and 15 metres but recorded an average error margin of 42.48% at 20 metres. It can be stated that for collision avoidance systems, a suite of sensors, including Ultrasonic, LiDAR and RADAR sensors, can effectively detect objects in their path with an accuracy above 97% without the use of vision sensors.

Abstract: Financial fraud remains a persistent challenge across various domains, particularly in public-sector financial operations, threatening the integrity and transparency of financial statements while eroding public trust. This highlights the need for continued advancement of fraud detection mechanisms to keep up with the ever-evolving fraud tactics. ML algorithms have proven to be one of the most successful methods for analysing large financial datasets to detect fraudulent patterns. This paper reviews the application of ML to detect fraud in financial transactions using ML-based algorithms, namely K-means, Support Vector Machine, Decision Trees, Naive Bayes, and Deep Learning, in fraud detection, analysing their use cases and effectiveness as reported in the literature. Additionally, the study experimentally compares the performance of a Convolutional Neural Network (CNN) model against a Logistic Regression model, with the CNN achieving an impressive 90% accuracy, outperforming Logistic Regression in fraud detection. The paper further investigates the financial features and indicators most relevant to fraud detection and explores the challenges and opportunities posed by large volumes of financial transactions. By addressing these areas, the study aims to provide insights into enhancing fraud detection mechanisms and strengthening the security and integrity of financial transactions in today's digital ecosystem, including government institutions.

Abstract: There are a set of few among the 7billion+ people in the world with either hearing or speech impairment, their only means of communication is through the use of sign language. It is one of the most reliable methods of communicating with special needs people. A form of communication using visual patterns to express emotions and ideas helps bridge the gap for deaf individuals. However, when interacting with those who rely on spoken language, a communication barrier often arises. Currently, human interpreters are used to facilitate conversations between these groups, but this solution can be both costly and inconvenient. The necessity for developing a technology to aid the interpretation of sign language to the deaf community and to foster idea-sharing amongst all humans cannot be overemphasized. Much research has been carried out to acknowledge sign language using technology for most global languages. In this project, deep learning techniques were applied to develop a system for recognizing hand gestures in American Sign Language. A dataset was created using both two-dimensional and three-dimensional images of American gestures. To detect landmarks in these images, the MediaPipe framework was utilized. Additionally, Long Short-Term Memory (LSTM) networks were employed to improve gesture recognition by leveraging the temporal dependencies of hand movements. My work specifically focuses on recognizing contexts in sign language communication, enhancing the system's ability to understand not just individual gestures but also their meanings in different scenarios.