I am a passionate seeker of knowledge dedicated to the continuous elevation of human civilization through advanced engineering design and technology.
Office Hours
My Journey
Autonomous Vehicles Laboratory
Toronto, ON Sep 2023 ~
MASc, Aerospace Engineering
My research aims to revolutionize the field of Uncrewed Aerial Vehicle (UAV) detection by generating a large and diverse dataset specifically designed for vision-based air-to-air detection scenarios. My approach leverages cutting-edge photorealistic simulation environments. By utilizing robotics simulation and game development engines, I can create a myriad of scenarios, encompassing various lighting, weather conditions, and UAV types, all while overcoming the challenges of traditional dataset generation methods.
To bridge the realism gap, my methodology incorporates domain randomization techniques, introducing crucial variations in environmental factors, nearby UAV characteristics, and camera parameters. This meticulous process ensures the creation of a high-variance dataset, essential for training Machine Learning (ML) models robust enough to handle diverse real-world conditions. The development of a Python script automates the introduction of random changes during simulations, resulting in hundreds of unique configurations for data acquisition.
The quality evaluation of my dataset involves training benchmark ML object detectors, including renowned models like RetinaNet, YOLO, and Faster R-CNN. Field experiments with a diverse set of UAVs complement the analysis, guiding the enhancement of my dataset through an iterative training and testing cycle. The ultimate goal is to establish a reliable Simulation-to-Reality (Sim2Real) framework for vision-based air-to-air drone detection.
By sharing this dataset with the public, I aim to propel the Urban Air Mobility field by advancing detection capabilities, contributing to the development of reliable algorithms crucial for enhancing DAA systems. The impact of our research extends to facilitating BVLOS operations, unlocking the full potential of UAVs to positively impact the lives of Canadians.
MetRocketry
Toronto, ON Sep 2022 - Apr 2023
Aerostructures Engineering Assistant
As a member of the MetRocketry design team at my university, I played an instrumental role in the design, simulation, and manufacturing of composite rocket fins for our two-stage rocket. The fins were made from 12 plies of laminated woven carbon fiber, and I conducted a flight profile simulation using OpenRocket to obtain key design parameters such as max velocity, max altitude, and maximum dynamic pressure. From these parameters, I set up a coupled simulation in ANSYS Workbench to run an aeroelastic analysis on the fin designs for each stage of the rocket. By doing so, I was able to determine whether aeroelastic flutter would occur in our expected flight envelope.
My work on the MetRocketry design team was critical in ensuring the success of our rocket's flight performance. Through my simulation and analysis work, we were able to identify and prevent aeroelastic flutter, which could have resulted in catastrophic failure during flight. The importance of this work extends beyond the design and construction of the rocket itself, as it highlights the importance of thorough and accurate analysis in engineering projects. As I move forward in my career, I will continue to apply the lessons I learned during my time with the MetRocketry team and prioritize the importance of meticulous analysis in all of my engineering projects.
Maple Advanced Robotics Inc.
Richmond Hill, ON May 2022 - Sep 2022
Research and Development Engineering Intern
In this role I had the opportunity to showcase my proficiency in MATLAB programming to a highly educated multidisciplinary team of 18 engineers to aid in the research and development of the start-up company's proprietary robotic tool path planning technology. Being the only undergraduate student still on my way to completing my first degree, in an office otherwise occupied only by engineers with PhDs and many years of research and development experience in their respective fields of robotics, computer vision and advanced mathematics was daunting at first. However, this put me in a unique position to learn and benefit greatly from my peers' and supervisors' wealth of knowledge and experience, specifically in the implementation of new technologies in industries that demand it. While the bulk of my previous programming experience has been attached to solving my own mathematical models formulated for specific problems and assignments, or for automating repetitive tasks, my most valuable takeaway from this experience was that I began to learn about different optimization techniques necessary to make the most efficient use of computational resources when designing algorithms.
Although much of the project specifics are beyond the scope of what I'm allowed to discuss, my specific role and key contributions to the team include the development of an algorithm which takes a large point cloud matrix often consisting of more 10,000-point locations in 3D space which describe the surface geometry of industrial workpieces with mm accuracy as input, to then output a polygonal surface projection defined by a minimal number of points in 2D. I then developed a shape recognition algorithm which computes and stores the geometric properties of each scanned workpiece to workpiece profile. This was done to minimize the required user input for the operation of the company's tool path planning software, and to allow users to save path planning parameters to specific workpieces for ease of repeatability. The algorithms I developed were prototyped in MATLAB and went through many iterations to make the algorithm as efficient as possible before eventually being converted to a C++ package for implementation in the company's proprietary software. This presented a challenge for me as I was restricted in my use of MATLAB specific toolboxes and functions for this reason, but it forced to think outside the box and come up with ways to build my functions from a purely numerical standpoint and use matrix operations efficiently to maximize the amount of information I can store while minimizing the number of steps and calculations during the run sequence.
TMU Research and Innovation
Toronto, ON Feb 2022 - Sep 2022
Undergraduate Research Assistant
In this collaborative effort between the Toronto Metropolitan University and Bombardier Aerospace, graduate level researchers are investigating a multitude of aircraft cabin interior technologies as well as performing original studies on human factors which influence comfort to revolutionize the future designs of aircraft cabin interiors with a focus on smart technologies designed specifically for optimizing passenger comfort levels and reducing fatigue associated with air travel. Under the supervision of Dr. Reza Faieghi and Dr. Fengfeng (Jeff) Xi, I was a pivotal member of a team of graduate students conducting human comfort studies alongside OpenSim biomechanical simulations to investigate the human body-seat interaction. We investigated several comfort parameters including pressure distribution and electromyographic (EMG) readings of muscular activation in various seated positions and performed biomechanical simulations alongside our experimental trials and compared the measured experimental parameters with those produced by the OpenSim biomechanical simulation software for correlation. This was a pilot study which to the best of our knowledge is the first to use this method of biomechanical simulation for the purpose of optimizing the digital design of aircraft seats for human comfort.
I provided the team with an algorithm which I developed in MATLAB to automate several necessary data processing steps in the experimental procedure which we carried out on 20 subjects. Among the processes I automated which resulted in the most significant reduction of data processing and analysis time is an algorithm which reads raw experimental data directly from a pressure sensor mat in order to map the pressure distribution as external forces on the biomechanical model which has been scaled for each subject by their height and body mass and produces all the files in their unique formatting structure specific to the OpenSim simulation software. This drastically reduced the time it took between the experimental data acquisition part of the process and the simulation and analysis part. Beyond this I developed another algorithm to read the structured output files of the simulation to automatically generate the relevant plots and extract parameters of interest in this study to be stored in a relational database I had set up. The Access database was then linked to an Excel spreadsheet which contained functions to automate the statistical analysis techniques used on our dataset as its being updated. This allowed for a full streamlined workflow so that every effort was made to reduce the time required after experimental data acquisition to really analyze the quality of the data we were measuring. The journal article I co-authored is in the review process for publication.
Quadri Pharmaceuticals LLC
Dubai, UAE Aug 2021 - Dec 2021
Project Engineering Intern
During the height of the COVID pandemic, I was struggling to get my hands on paid internship experience in the Aerospace field, let alone in Canada or the United States, I didn't let that deter me and I expanded my horizons and continued my search and make a last-ditch attempt at a leading biomedical supply company in the whole MENA region and they operate out of the UAE. This company supplied numerous medical institutions ranging from clinics to hospital networks with all kinds of specialty equipment and surgical instrumentation etc. The company had hired an Electrical Engineer by the name of Eng, Mohammad Al Dunbok 3 months prior to the start of my internship. Eng. Dunbok had a vision to expand the company's business model by developing the Medical Projects Department in which he proposed that there was no reason why as a medical equipment supplier, we couldn't also extend our reach into the projects market to build entire labs or renovate older hospital wings. I was hired on to assist this engineer in forming the foundations for the new Medical Projects department at this company.
While initially thoughts of the distance from my family as well as the fact the field would be more biomedical than anything else, I knew that this was an opportunity for learning and that if I just take that step there can be a lot to gain in transferrable skills that I can carry with me in the future as well as a new perspective on a country I've never been to before. I took that chance and learned a lot of valuable lessons and made a great friend out of my mentor who I keep in contact with today.
During my time there I drafted a database of Microsoft Excel Templates for the medical projects to efficiently plan and implement projects, these templates included Risk Mitigation Charts, Contact Follow Up Sheets as well as technical datasheets for each part we supplied organized by supplier. I was fortunate enough to personally supervise the completion of the 2 turnkey projects in which I regularly reported project updates to a cross-functional team of engineers who represented SEHA, the Abu Dhabi Public Health Services Agency. Before leaving I prepared lengthy documentation and a presentation which form the standard processes and procedures for future hires to refer to as a means of ensuring the same quality of reporting and analytics are maintained in my leave.
Ryerson International Hyperloop
Toronto, ON Oct 2019 - Jan 2022
Structures Design Team Leader
I began my role at this team as a member of the chassis sub-team, our design objective was to reduce the chassis' empty weight and while maintaining lateral and longitudinal stability of the frame, which required strategic subsystem integration design. This project was for me a way to familiarize myself a little more with the iterative design process, and also taught me the essence of structure determines function. I also used this project as a way to practice my 3D CAD design skills with CATIA V5 as well how to set up FEA simulations in ANSYS Mechanical for simple loading cases in order to identify points of failure in the design. We achieved a design that's 40% lighter than our previous iteration while maintaining a similar power output from the linear induction motor which generated the pod's thrust.
In January of 2020 I was asked to assume a position of leadership to spearhead the efforts of our team to begin assembling our pod together, this role change came with added responsibility to the team, as team leader I became responsible for maintaining open channels of communication between the 4 structures sub-teams (Chassis, VDC, GNC, Propulsion) and was the primary decision maker for the integration of each of these teams' designs onto the chassis. I met some of my best friends in my university career during this time and will cherish all the times we spent in the lab together however frustrating it might've been sometimes in the thick of it.
Toronto Metropolitan University
Formerly Ryerson University Toronto, ON Sep 2017 - Apr 2023
BEng, Aerospace Engineering
Ever since I was a kid, my mom would always warn guests not to ask me what time it is, because I would respond with a 20-minute lecture on how the earliest clocks were invented. I never knew just how significant these little nuances our parents see in us early on until I really had to think about what I wanted to do with my life after high school. I was always fascinated by how the simple components can come together to form complex machines in the hands of those who take the time to develop a thorough understanding of the physical phenomena at play. Throughout my whole life I've grown a profound passion for observing the wonders of the world we live in which then developed into an obsession for peeling back the curtains and trying to develop an understanding of how all these systems come together. This obsession very quickly fixated my eyes on the skies, and I set out a goal for myself that I need to do what I can to reach the heightened level of understanding that humanity needed to turn metal pieces into flying machines which totally changed humanities perception of distance.
I've finally earned by Bachelors of Engineering in the field of Aerospace Engineering and every day I thank God first and my parents second for giving up so much and overcoming the greatest of obstacles so that I could focus on building my wings (pun intended) and pursue my passion and I never think about what I would have studied if it hadn't been Aerospace because it feels like there was no better suited program for with my unique perspective and passion. My lifelong goal is to be able to continue to lead a life full of learning and sharing what I've learned with the world by developing new sustainable technologies needed to overcome the world's present obstacles while also continuing to pursue my passion of teaching as a professor in a classroom to share my passion with like-minded students to do my part to give back and develop the next generation of bright youth the way my many teachers have done for me.
Royal Pro Inc
Mississauga, ON 2013 - Always
The Boss's Son
Like most children of immigrant parents, I learned the meaning of hard work from my father. In my younger days I got to witness firsthand what my father was sacrificing every day for us and so whenever we had time off from school, he made sure we were familiar with the weight of that sacrifice, often times that weight came in the form of heavy drywall sheets. My father is always fair and honest with everyone we work with, and that includes his sons, but when you're the boss's son, like the boss you get to wear many different types of construction hats. I've become a jack of all trades, decently skilled in the arts of framing, and flooring and drywall, a master of none of course but decent, nevertheless.
As I grew older my ability to "pull weight" also increased, I figured as my father is investing himself into this business for our academic development, the least I can do is share with him the fruits of his labors by reinvesting some of the skills I've picked up elsewhere back into my father's business. With my newfound Excel expertise, I began tracking all my father's business expenses, I developed spreadsheets for every project and tracked every cost with indicators for material or labor costs as well notes to keep track of which stage of the project each cost arose. I also built handy spreadsheet tools he can take on the go when drafting quotations and used this information to reference the real costs during the project. The addition of these data analytics and insights allowed us to efficiently track in real time our costs in reference to our budget and minimize cost overrun to maximize take home profit. I currently manage several projects for my father's construction company with budgets ranging from $50,000 - $300,000. As far as this world has taken me, this job of mine has been the most intrinsically rewarding as it allows me the chance to repay only a fraction of what my parents have given us and that will always be my primary job responsibility regardless of where I'm employed.
At the essence of my graduate my research is the use of AirSim, a cutting edge flight-simulation tool originally developed by Microsoft which is hosted in Unreal Engine and showcased in this video. It's primary strength is in its ability to generates realistic views from the perspective of a multirotor, while providing simultaneous segmentation and depth images.
My research is focused on the use of domain randomization techniques to arrive a Sim2Real framework which I implement by developing Python scripts to interact with the AirSim API
Design of Cargo Freighter Aircraft
Nearing the end of my undergraduate studies, I had the opportunity to participate in a capstone design project supervised by senior Bombardier engineers. Our group was tasked with designing a transporter aircraft which would deliver the centre fuselage of the Global 7500 business jet from Wichita, KS to Toronto, ON. I assumed a position of leadership on my team and was personally responsible for the structural design of our airframe. This included developing algorithms to determine the appropriate sizing for structural members based on our load case scenarios, material selection, and even the 3D digital design in CATIA V5.
Our high-level configuration featured a high wing design, and T-Tail configuration, the loading and unloading of our cargo was faciliitated by a unique swing tail design, This project required me to develop detailed plans, conduct research and present our findings to senior engineers, which allowed me to develop my teamwork and communication skills which I believe are essential for any engineer.
Design of Supersonic Inlet
In this project, completed as part of my Aircraft Propulsion course, the requirement was to design a supersonic inlet which is adapted for a flight mach number of 2.4, and having a mach number of 0.79 for the flow entering the combustion chamber. The design was to include a shock system which consists of 3 oblique shocks followed by a normal shock. To solve this problem, I designed an algorithmic approach which took these input requirements and solved a system of shock equations to determine the necessary ramp angles which would produce the shocks that satisfy these requirements. To adapt the aerospike for the flight mach number, I designed another system of simpler trigonometric relations to determine the lengths of each ramp required to ensure that all shock cones met at a singular point which was coincident with the cowl lip to optimize the design for maximum intake capture area. I then scaled the design to match the mass inflow rate of the J58 Engines which powered the SR-71 Blackbird and performed a 1-Dimensional parametric analysis for both my designed inlet and that of the J58 engine to compare their propulsive performance characteristics.
Aeroelastic Simulation of Rocket Fin in ANSYS
This animation shows the contoured deformation of the composite rocket fin under the expected flight loads obtained from a coupled simulation in ANSYS Workbench. The simulation was set up by coupling the transient structural and fluent fluid flow solvers to run an aeroelastic analysis on the fin designs for each of the 2 stages of the rocket. The transient structural solver simulated the structural deformation of the fin due to aerodynamic loads, while the fluent solver calculated the unsteady aerodynamic loads acting on the fin. By combining the results of both solvers, we were able to determine if aeroelastic flutter would occur in our expected flight envelope. This analysis was crucial in ensuring the stability and safety of our rocket during flight.
OpenSim Biomechanical Simulation
This is a quick demonstration of how the OpenSim Biomechanical simulation was used for the purpose of research on the internal musculoskeletal forces experienced during the human body-seat interaction. Required simulation inputs include the model's height and mass for proportional scaling, and the individual joint positions as well as external forces applied on the model are encoded to output an approximated internal muscular activation percentage as well as a magnitude of internal muscular forces displayed for visual purposes as colored contours. The purpose of this study was to develop a method for assessing the comfort levels of various digital seat designs in order to optimize airline cabin interiors for comfort with the addition of human factors and analytics.
Algorithm to Obtain 2D Trace Outline of 3D Point Cloud Image
Without going into too much technical detail, the objective of this algorithm is essentially take the 3-Dimensional output of a laser scan of surface, whatever the workpiece might be, and process the 3-Dimensional array of points using Linear Algebra to produce a minimal set of 2-Dimensional points while maintaining unique geometric shape properties. The 2-Dimensional surface is a necessary input in the tool path planning process. To extend the functional use cases of this algorithm, I also developed a shape recognition algorithm which classifies workpieces by their unique geometric features so that previously designed tool path plan parameters can be recalled by the computer when it recognizes a part it's seen before.
Hyperloop Pod Structure Design
This animation was rendered to showcase our team's pod design assembly including all its major subsystem components. In the center rests the linear induction motor and the batteries were placed over top. This specific configuration called for an EMF shield which propped the batteries up at a calculated distance from the LIM to minimize electromagnetic induced loss of thrust. On either side of the LIM are the pneumatically actuated brake calipers, which are pressurized by 2 air tanks taken off of paintball guns.
Publications
Neural Moving Horizon Estimation: A Systematic Literature Review arXiv