Addise Endale Fentaw
Course level: Masters
Research title: Machine Learning-Based Predictive Control Strategy for Active Cell Balancing in EV Battery Packs
Executive Summary: Despite advancements in battery technology, a significant challenge in Electric Vehicle (EV) battery packs is cell imbalance, leading to reduced capacity, accelerated degradation, shortens lifespan, poses safety concerns and limit overall vehicle efficiency. Conventional battery management approaches, such as passive balancing and simple rule-based control, often do not adapt well to dynamic conditions and may waste energy or fail to extend battery life optimally. This study aims to address this fundamental problem by exploring advanced control strategies for active cell balancing. It will investigate how highly efficient, data-driven approaches, such as those combining machine learning with predictive control, can be applied to optimize charge distribution among battery cells. The ultimate goal is to develop a more efficient and responsive balancing system that significantly enhances battery longevity, boosts performance, and contributes to extending the operational range of electric vehicles.
Research Question: How can control strategies, incorporating machine learning and predictive techniques, be developed to optimize active cell balancing for improved performance, lifespan, and efficiency in Electric Vehicle (EV) battery packs?
Contact: addise.endale@students.jkuat.ac.ke
Awe Victoria Damilola
Course level: Ph.D
Research title: Predictive Modelling of Additive Manufacturing of PEKK for Electric Motorbike Motor Housings
Executive Summary: This research focuses on predictive modeling of additive manufacturing (AM) of CF-PEKK for electric motorbike motor housings. CF-PEKK, a high-performance thermoplastic composite, provides a better strength-to-weight ratio, thermal stability, and corrosion resistance than traditional materials. Using Fused Filament Fabrication (FFF), this study will develop optimized manufacturing parameters through DIGIMAT simulations and Artificial Neural Networks (ANNs) to improve mechanical performance. These include tensile strength, fatigue resistance, hardness, and residual stress. The fabricated motor housing will be tested and compared to conventional ones regarding performance and cost. The outcome will support lightweight, durable, and sustainable electric mobility solutions.
Research Question: How can AI-optimized Fused Filament Fabrication (FFF) of Carbon Fiber-Reinforced PEKK composites be used to develop lightweight, corrosion-resistant, and fatigue-resistant motor housings for electric vehicles, as an improvement over traditional metal-based designs?
Contact: awe.victoria@students.jkuat.ac.ke
Boniface Ntambara
Course level: Ph.D
Research title: Development of PCM-Based EV Battery Thermal Management Scheme with Real-Time Battery Monitoring and Temperature Estimation System
Executive Summary: The proposed research aims to develop an integrated, passive Battery Thermal Management System (BTMS) for electric motor bicycle that combines phase‐change materials (PCMs) with real‑time sensing and an AI‑driven temperature‑estimation model. A PCM‐based cooling assembly will be designed with an IoT based thermal monitoring system. A Recurrent Neural Network (TD‑RNN) will be developed and trained to predict battery state of health. This research will deliver a cost‑effective, scalable solution that enhances EV safety, longevity, and performance.
Research Question: How can the battery state of health be impacted by using passive battery thermal management?
Balogun Victoria Adebisi
Course level: Masters
Research title: Design and testing of battery placement for self-cooling for e-bikes
Executive Summary: The goal of this study is to relate efficiency and safety of the battery location for temperature control. The study would, therefore, involve development of a Computational Fluid Dynamics model for the battery pack positioning. The study will analyze the suitability of battery positioning for passenger health.
Research Question: How safe is battery location on an e-bike and how does it support cooling?
Contact: balogun.victoria@students.jkuat.ac.ke
Abrham Degarege
Course level: Masters
Research title: Numerical Investigation of PCM-Based Battery Thermal Management System
Executive Summary: This research project focuses on enhancing the thermal control of lithium-ion batteries through the integration of Phase Change Materials (PCMs). During operation batteries produce heat that if left unmanaged, it can negatively impact their safety, efficiency, and overall longevity. This study investigates the potential of PCMs, which absorb and release thermal energy during phase transitions, to regulate battery temperature more effectively. The research involves developing a thermoelectric model to evaluate the behavior of a PCM-based battery system under realistic operating conditions. In addition, it will analyze
impact of thermal and electrical variations on overall performance.
Research Question: How effective are Phase Change Materials (PCMs) in improving the thermal management of lithium-ion battery systems as evaluated through numerical simulation?
Contact: alemu.abrham@students.jkuat.ac.ke
Boesi Phetogo
Course level: Masters
Research title: Topology optimization for carbon fibre storage for custom EV bikes
Executive Summary:This study involves the design, simulation, and fabrication of a carbon fibre storage unit for electric bikes. Topology optimization techniques will be employed to determine the most efficient material layout based on performance requirements and given constraints. This approach will result in a high strength-to-weight ratio for the product. Based on the final optimized design, the storage unit will be fabricated using 3D printing and tested for relevant mechanical properties and functionality.
Research Question: How can topology optimization be applied to design a lightweight yet structurally efficient carbon fibre storage unit for custom electric bikes?
Contact: phetogo.boesi@students.jkuat.ac.ke
Eng. Geoffrey Hoseah Okoth
Course level: Ph.D
Research title: Selective Laser Melting of A2024-RAM5 Open Differential Gears for Electric Vehicles
Executive Summary: This study focuses on open differential EV gears which are often
manufactured using conventional machining processes. These gears are prone to excessive
wear, plastic deformation, and heat-induced pitting, causing gear damage. This study aims to evaluate the
mechanical performance of A2024 open differential spider gears for
electric vehicles manufactured using selective laser melting. SLM will be combined with DfAM to fabricate topology-optimized open
differential gears for EVs through material substitution. Laser-based
manufacturing techniques are attractive because they provide
precise control which maximizes material use,
allowing for zero-defect and zero-waste parts production.
Research Question: How does SLM
combined with DfAM improve the mechanical performance, wear resistance, and
structural efficiency of topology-optimized A2024-RAM5 open differential spider
gears for electric vehicles compared to conventionally manufactured gears?
Contact: geoffreyhoseah@gmail.com
Talent Kachomba
Course level: Ph.D
Research title: Laser fabrication of lightweight and high-strength functionally graded rotor shaft for e-mobility
Executive Summary: The transition to EVs introduces both challenges and opportunities, particularly in the development of advanced propulsion system components such as rotor shafts. As a critical element accounting for over 30% of the rotor’s weight, the rotor shaft directly influences motor efficiency, torque delivery, and energy consumption especially during high-demand scenarios like vehicle acceleration. Rotor shaft design must balance high performance, lightweight construction, effective cooling, and robust reliability, while remaining cost-effective and sustainable. This study uses laser-based manufacturing for EV rotor shaft development using functionally graded materials. Topology optimization and machine learning is used to achieve lightweight and robust designs, and to optimize processing conditions. The proposed bM framework will enable the fabrication of FG rotor shafts with tailored material properties for high-performance and reliable EV propulsion systems.
Research Question: How can the integration of laser-based manufacturing (LbM), functionally graded materials (FGMs), topology optimization (TO), and machine learning (ML) be leveraged to design and fabricate high-performance, lightweight, and reliable rotor shafts with tailored material properties for next-generation electric vehicle (EV) propulsion systems?
Contact: talentkachomba3@gmail.com
Natnael Abebe Tsega
Course level: Masters
Research title: Machine Learning Driven Process Optimization of Laser AM Light weight alloy for EV Motor Housings: Achieving Synergistic Thermal and Mechanical Performance
Executive Summary: When applied to lightweight alloys, SLM presents notable challenges including process-induced defects, residual stresses, and anisotropic microstructures that compromise component performance. Furthermore, the relationship between thermal conductivity and microstructure in as-built SLM alloys remains poorly understood. This project introduces an innovative framework combining a hybrid Physics-Informed Neural Networks (PINNs + Bayesian networks) and NSGA-III multi-objective optimization. PINNs will integrate sparse experimental data with underlying thermo-mechanical principles to model microstructure-property relationships and Bayesian networks predicts defect formations. NSGA-III will identify Pareto-optimal LPBF process parameters. These tools will advance the understanding of interactions among laser process variables, alloy composition, and resultant mechanical and thermal properties in EV motor housing representative test coupons.
Research Question: How can multi-objective machine learning optimization reconcile the potential trade-offs between thermal conductivity and mechanical strength in laser AM processed lightweight alloys specifically for EV motor housing applications?
Contact: natnaelabebe05@gmail.com
Thomas Francis Memirieki
Course level: Masters
Research title: Machine Learning-based Laser Metal Deposition Parameters Optimization of the Functionally Graded 300 Maraging steel/SiC Electric Vehicles Rotor Shaft
Executive Summary: The study aims to optimize the LMD parameters for manufacturing FGMs 300 MS reinforced with SiC, tailored for rotor shaft applications in EVs. The research focuses on developing a lightweight and enhanced mechanical properties for greater efficiency and long-distance driving. This work applies the appropriate ML that can deal with complex relationships between the most influential deposition parameters such as laser power, scan speed among others on the responses (porosity, density and microhardness). Minitab and Experimental data from the multi-track and multi-layer depositions will be used to train, validate and test the predictive model. The research will not only end with fabricating a robust EV rotor shaft but also open the door for different industrial applications.
Research Question: How can ML be applied to optimize the LMD parameters to achieve defect-free and high-performance 300 MS/SiC FGM rotor shaft for EVs?
Contact: mt25020300@biust.bw.ac
Paul Mupenzi
Course level: Masters
Research title: Process Parameter Optimization of Laser-Engineered Net Shaped Functionally Graded Composites for Electric Vehicle Spider Gears
Executive Summary:This study will focus on optimizing Laser-Engineered Net Shaping (LENS) process parameters in functionally graded 20MnCr5-TiC-SiC composites for electric vehicle spider gears. It will experimentally investigate how LENS process parameters affect microstructure and mechanical properties. Machine learning techniques will be used to optimize these parameters with the goal of minimizing porosity and improving hardness and wear resistance. The outcome will be a validated, data-driven process for producing defect-free, high-performance spider gears for EV applications.
Research Question: How can machine learning optimize LENS parameters to minimize porosity and improve mechanical properties (hardness, wear resistance) in 20MnCr5-TiC-SiC functionally graded composites for high-performance EV spider gears?
Contact: mp25020284@biust.ac.bw
Sheriff Olalekan Saka
Course level: PhD
Research title: Simulation and Modelling of Microstructure and Mechanical Properties of 3D-Printed Metallic Alloys and Composites for Electric Vehicle Applications
Executive Summary:This study aims to perform simulation and modelling of microstructure and mechanical properties of three-dimensional (3D) printed metallic alloys and composites intended for electric motor housing applications in EVs. Al-10Si-Mg will be utilized because of its additive manufacturing (AM) suitability, strength-to-weight ratios, and thermal behaviour. Laser Engineered Near Shaping (LENS) and Selective Laser Melting (SLM) methods will be used due to their capability to fabricate intricate geometries and high-performance metallic alloys. The multi-scale modelling approaches to be employed will enhance the quantitatively the refinement of grains, phase stability, and porosity. All these quantified data are essential for mechanical behaviour such as fatigue resistance and physical property like thermal conductivity. The outcomes of the study will aid the understanding and application of rapid prototyping, and contributes to sustainable application through advanced materials engineering.
Research Question: What are the significances of multi-phase simulations and modelling for the predictions of microstructure evolution and mechanical properties of newly 3D-printed metallic alloys and composites for EV applications?
Contact: ss24020236@biust.ac.bw
Jegede Oluwabusuyi Samuel
Course level: Masters
Research title: Laser-Fabricated Electric Vehicle Motor Housings: Integrating Materials Design with Process Optimization
Executive Summary: This project investigates the use of laser-based manufacturing specifically laser powder bed fusion and laser welding for fabricating motor housing components from advanced composites or aluminium alloys. The chosen materials are expected to provide high strength-to-weight ratios, good thermal conductivity, and impact resistance. The study will focus on adjusting laser parameters for minimal porosity and optimized grain refinement, leading to enhanced mechanical and thermal properties. Performance evaluation will involve mechanical testing (e.g., impact, tensile, and fatigue tests) alongside thermal conductivity measurements and vibration analysis. Additionally, finite element analysis (FEA) will simulate thermal and mechanical stress distributions, guiding further process optimization. The outcome is anticipated to yield a motor housing that meets the demands of next-generation e-mobility systems, paving the way for more efficient, reliable, and lightweight electric vehicles.
Research Question: What materials design strategies can be employed during laser processing to enhance energy efficiency and impact resistance in motor housing components?
Contact: jegedeoluwabusuyisamuel@gmail.com
Emmanuel Efemena Lindsay
Course level: Ph.D
Research title: Powder Assisted Laser Beam Welding process optimization for lightweighting of AA6063-T6 EV structural components
Executive Summary: The study will involve powder (Ti6Al4V) assisted Laser Beam Welding process optimization of aluminium alloy 6063-T6 for EV components. Process optimization will be done with the aim of minimizing porosity while maximizing joint strength. Results will be compared to conventional LBW methods currently used to join EV body- or structural panels. A set of experiments (in line with Machine Learning techniques) will be performed to develop an algorithm that can predict optimum process parameters for maximum strength.
Research Question: Can Machine Learning techniques be used to successfully predict process parameters to obtain maximum mechanical properties i.e., strength, hardness, fatigue, for AA6063-T6?
Contact: s232141533@mandela.ac.za
Leila Mbagaya
Course level: Ph.D
Research title: Machine Learning techniques for battery health prediction
Executive Summary: The aim of the project is to provide an in-depth review of the performance of the various machine learning (ML) techniques with regards to State of Health (SoH) prediction for batteries. The study will identify the various parameters relative to such prediction, and critically evaluate the efficacies of numerous ML techniques. An existing ML technique will be modified, which results in an increase in prediction reliability. Further, the use of metaheuristic optimization techniques will be investigated, both for the optimization of ML specific parameters, and for feature selection. A prototype of the battery SoH monitoring system will be designed and constructed.
Research Question: What are the prevailing and emerging machine learning approaches for forecasting battery State of Charge (SoC) and State of Health (SoH), how do these methods compare in terms of performance, and how do factors such as relevant features, metaheuristic techniques, and hybrid models influence prediction accuracy?
Contact: s232216886@mandela.ac.za
Christine Mutile
Course level: Masters
Research title: Laser beam Welding Process Optimization of thin stainless steel sheet AISI304L (0.5mm)
Executive Summary: The study will involve process development and optimization of laser beam welding of thin stainless steel sheet. A laser beam welding process will be developed for optimized joint integrity for both keyhole- and conduction mode LBW. The laser beam welding process will include both keyhole- and conduction mode. The research design will ensure generation of data sets that can be used in applying ML techniques for process prediction.
Research Question: Which process characteristics are the main contributors to obtain acceptable joint integrity?
Contact: s232280258@mandela.ac.za
Mr Michael Femi Olabisi
Course level: Masters
Research title: Laser-based heat treatment process development for Laser Metal Deposition (clad) layer/s on heat sensitive alloy 17-4PH
Executive Summary: The study will involve the development of a laser-based heat treatment process to restore the properties of heat sensitive alloy 17-4PH which have been degraded due to heat intensive processes such as Laser Metal Deposition (LMD/DED), welding, etc. Heat sensitive alloys are commonly used for high strength applications and often components need to be replaced as the alloy does not allow heat intensive refurbishment processes such as directed-energy-deposition (DED) which is ideal for the refurbishment of high value components.
Research Question: This project will investigate the viability (based on mechanical properties and process cost) of employing a laser-based heat treatment process for the restoration of parent metal properties.
Contact: s232271240@mandela.ac.za
Laone Ronaldo Pilane
Course level: Masters
Research title: Machine Learning Based Optimisation of Laser Metal Deposition Parameters for Al-Si Alloy Electric Vehicle Component Refurbishment
Executive Summary:The success of LMD is particularly important for repair applications where dimensional accuracy and structural integrity are critical and depend on precise parameter optimisation. Traditional optimisation techniques such as Design of Experiments (DoE) and Response Surface Methodology (RSM) have been extensively used to relate process parameters to output quality. Unfortunately, these statistical trial-and-error methods can be time consuming and mostly material-specific, making them less flexible for broader applications. Studies on aluminium-based alloys, which are structurally and mechanically different, remain unexplored. Therefore, this research project aims to provide valuable additions to the body of knowledge in this field.
Research Question: To develop a machine learning-based framework for optimising LMD process parameters to enhance the repair quality and mechanical performance (based on microhardness) of cast aluminium-silicon alloy (LM27) for electric vehicle components
Contact: s232273200@mandela.ac.za
Admire Chityori
Course level: Ph.D
Research title: Digital modelling for Laser Directed Energy Deposition Manufacturing process of light weight electric vehicle components
Executive Summary: This study aims to develop a digital twin model to streamline the Laser Directed Energy Deposition (LDED) production process for next-generation lightweight electric vehicle components. The digital model will replicate the physical LDED system, enabling “what-if” simulations, process parameter (laser power, scan speed) optimization, and real-time monitoring and control. In-situ melt-pool thermal data will feed into a machine-learning based model-predictive control (MPC) system that dynamically adjusts laser power and scanning speed. This closed-loop system will detect and minimise key defects (keyhole porosity, lack-of-fusion porosity, and residual stresses), during fabrication, ensuring “first-time-right” manufacturing principles for complex EV components. By integrating a Product, Process, and Resource (PPR) methodology within the digital twin framework, the project seeks to reduce lead times, enhance build quality, and lower production costs. The outcome will be a comprehensive PPR-driven digital-twin platform capable of adaptive control of LDED for high-precision EV components.
Research Question: How can a digital twin integrating real-time melt-pool monitoring, machine learning-based model predictive control, enhanced by a PPR framework, optimize LDED process parameters to minimize process-induced defects (keyhole porosity, lack of fusion porosity, residual stresses) and achieve first-time-right manufacturing of next-generation EV components?
Contact: dope250005@pentvars.edu.gh
Amos Magari Nyakundi
Course level: Ph.D
Research title: Predictive modelling of thermal distortion of laser additive manufacturing of electric vehicle motor housing
Executive Summary:The aim of the project is to develop a predictive modelling framework to predict thermal distortion in laser-based manufacturing (LbM) of electric vehicle motor housing. The study will use Physics-Informed Machine Learning (PIML) to accurately predict thermal distortion in LbM. The model will incorporate key process parameters such as laser power and scan speed and their influence will be investigated. Further, process optimization will be conducted to improve part quality and reduce component failures.
Research Question: How can predictive modelling techniques be developed to estimate thermal distortion in LbM of electric vehicle motor housings, considering process parameters and material behavior?
Contact: dope250002@pentvars.edu.gh
Elphas Kibet Tum
Course level: PhD
Research title: Lifecycle Engineering of Laser based Manufacturing Systems for EV Motor Housing Manufacture
Executive Summary: This project aims to conduct a comprehensive Life Cycle Assessment (LCA) of additively manufactured (AM) electric vehicle (EV) motor housings, with a focus on the production phase. Unlike traditional component-based LCA approaches, this study treats the additive manufacturing process itself as a product. It evaluates the sustainability performance of the AM process by analyzing energy consumption, material efficiency, and potential for carbon footprint reduction. The goal is to generate actionable insights that can inform the sustainable design and manufacturing of next generation EV components. The outcomes will support industry efforts to reduce environmental impacts while ensuring structural integrity and long-term durability of critical vehicle parts.
Research Question: What are the environmental impacts associated with laser based additive manufacturing of EV motor housings, and how can lifecycle engineering approaches be used to improve process sustainability?
Contact: dope250003@pentvars.edu.gh
Oyeleke Abduljabar Gbenga
Course level: Masters
Research title: Lifecycle Assessment of laser welding systems for electric vehicle battery pack assembly
Executive Summary: This project investigates the environmental impacts of laser welding processes used in electric vehicle (EV) battery module assembly, with a focus on identifying key environmental hotspots across the product’s life cycle, production (gate) and end-of-life (EOL). While laser welding is favored for its precision and efficiency, its sustainability profile remains underexplored, especially in terms of energy consumption, emissions, and waste generation. The study employs Lifecycle Assessment (LCA) impact assessment method to evaluate critical parameters. These include material use, energy demand during welding operations, emissions from power consumption, and waste outcomes during disposal or recovery. The gate and EOL approach helps assess both the operational and disposal phases of laser-welded battery modules. By identifying the major contributors to environmental burdens (hotspots), this research will inform process optimization strategies and sustainable design practices. The findings aim to support zero-defect and zero-waste manufacturing initiatives, contributing to greener and more responsible EV production within the framework of Industry 4.0.
Research Question: How does laser welding in EV battery pack assembly impact the environment across the gate and end-of-life stages, and what are the key contributors to its overall environmental footprint?
Contact: mies250001@pentvars.edu.gh
Omotoso Samuel Oluwasina
Course level: Masters
Research title: Review and Selection of Alternative Manufacturing Processes for Sustainable Electric Vehicle Production
Executive Summary: The main aim of the study is to use LCA methodologies to identify the most sustainable and cost-effective manufacturing process for EVs. Alternative manufacturing methods will be compared with laser-based manufacturing methods to justify the economic, environmental, and sustainability performance.
Research Question: Which is the most sustainable and cost-effective manufacturing processes for EV production?
Contact: omotososamueloluwasina@outlook.com
Fabrice Ganza
Course level: Masters
Research title: Digital modeling of laser welding systems for battery packs assembly in Electric vehicles
Executive Summary: The study will involve development and validation of a digital simulation framework to improve laser welding processes in EV battery packs assembly. The study will focus on modelling the thermal, mechanical, and metallurgical behavior of materials during welding, using finite element and multi-physics techniques. Heat input, material deformation, and microstructural evolution will be simulated under various welding parameters, followed by experimental validation in a controlled laboratory environment. The work will also explore optimal welding conditions to minimize defects and enhance joint strength, considering the complex, multi-material nature of battery packs.
Research Question: How can digital modelling techniques be used to accurately predict and optimize the thermal, mechanical, and metallurgical behaviour of materials during laser welding, in order to improve the quality, reliability, and efficiency of battery pack assembly in electric vehicles?
Contact: fabriceganza2@gmail.com
Anteneh Misikir Melkamu
Course level: Masters
Research title: Predictive modelling of laser welding defects in EV battery pack Assembly
Executive Summary: This project focuses on developing predictive modelling of laser welding defects in electric vehicle (EV) battery pack assembly, to improve weld quality, reliability, and process efficiency. Laser welding defects such as porosity, cracks, lack of fusion, and spatter can compromise battery performance and safety. Machine learning (ML) techniques are used to develop data-driven models that predict the occurrence of defects based on key input parameters including laser power, welding speed, focus position, and material properties. By integrating real-time sensor data such as thermal images, acoustic signals, and optical monitoring with historical quality inspection results, the models can detect and classify defects with high accuracy. The proposed predictive model enables proactive quality control, minimizes post-weld inspection costs, and supports adaptive process optimization. This approach advances smart manufacturing for EV battery packs, aligning with Industry 4.0 goals and contributing to more durable, cost-effective, and safer electric vehicles.
Research Question: Can machine learning able to develop predictive model to minimize laser welding defects in EV battery pack assembly especially for battery tabs, bus bars and battery casings?
Contact: mies@pentvars.edu.gh
