»Home»Capstone Design Team Showcase: Tire Pressure Monitoring and Computer-Assisted Surgery Projects
Capstone Design Team Showcase: Tire Pressure Monitoring and Computer-Assisted Surgery Projects
September 13, 2024
Capstone design projects are a major component of the Department of Electrical and Computer Engineering curriculum. Students design a product or service of significance and solve an open-ended problem in their field of study.
ECE hosted a project showcase for the summer term on August 8th, 2024.
Wirelessly-Powered Tire Pressure Monitoring System
Ardavan Pourkeramati, Mihir Nimgade, Nick Zhao, Ethan Joyce Partner: UBC ECE Communications Lab and Sierra Wireless
Project and Solution
A common problem in the trucking industry is the need to remap tire pressure sensors after tire swaps. Without remapping, the system is unable to determine which pressure readings originate from which tire. This remapping requires manual technician labor and is an additional expense for trucking companies. Our team has developed a battery-free, wirelessly powered, Tire Pressure Monitoring System (TPMS) capable of performing this remapping automatically. Our system is able to selectively power the individual sensors in each tire, allowing it to always know which tire each pressure reading originates from, eliminating the need for manual remapping.
Challenges
During this project, we encountered several challenges that pushed us to expand our engineering skills. One of the key challenges was applying the theoretical Radio-frequency (RF) knowledge we learned in class to real-world scenarios. This required hands-on experience with testing and debugging using specialized RF equipment such as spectrum analyzers and VNAs. We also had to dive deep into researching and understanding RF energy harvesting, a field that is intricate and full of complexities. Identifying the most suitable harvesting solution took considerable time and experimentation to ensure it met our specific needs.
Designing the RF printed circuit boards (PCBs) presented its own set of difficulties. Our project required two PCBs–one for transmitting and one for receiving RF power. This process wasn’t just about creating a functional design; it demanded extensive research and testing to effectively integrate RF principles with embedded systems for optimal performance.
Future Impact
We hope that this project has successfully demonstrated an alternative method for performing automatic tire location mapping (i.e. “auto-localization”) for large vehicles such as tractor-trailers. Our solution favours a hardware-based approach by using RF energy harvesting (RFEH) technology to individually power on the embedded tire sensors. For our system, it was the application of RFEH that made the auto-localization problem tractable.
Computer Vision Processing for High-Resolution Angle Sensor in Computer-Assisted Surgery
Eddy Nangia, Michael Perkins, Charlotte Luo, Yifeng Liu, Zelin Li Partner: ISTAR Group, Vancouver General Hospital, and UBC Department of Mechanical Engineering
Project and Solution
The ISTAR group at the University of British Columbia is pioneering a new surgical system for mandibular reconstruction surgery. This system integrates mechanical devices with intraoperative surgical guidance to improve surgical outcomes. A critical aspect of this system is accurately tracking the positions of objects in the surgical field, such as mechanical devices and bone segments, using optical markers. However, the current optical markers are large and often obstructed, leading to suboptimal performance.
Our team’s objective was to replace these markers with a newly developed system called LentiMark. When used with a standard camera, LentiMarks can accurately measure position and orientation while being smaller and less obtrusive. This innovation aims to enhance ISTAR’s surgical guidance system, resulting in more effective operations.
The project focuses on developing and implementing a high-resolution angle sensor for computer-assisted surgery, using advanced computer vision techniques to improve surgical precision and reliability. The core innovation is the LentiMark, a novel optical marker combining ArUco markers with Variable Moiré Patterns (VMPs). While ArUco markers provide preliminary pose information, VMPs offer greater angle detection accuracy by varying visual effects with the viewing angle. Experimental results show that angle measurements maintain an error margin within 1 degree, significantly enhancing the accuracy and efficiency of surgical tools and thereby benefiting surgical navigation.
Challenges
Our team members come from diverse backgrounds, each with different strengths. Effective teamwork allowed us to support each other throughout the term, overcoming challenges and learning together. One significant challenge was the initial research, as some aspects of the project involved medical and optical knowledge not directly related to electrical or computer engineering. The design of Variable Moiré Patterns (VMPs) was particularly challenging due to the limited research available, so we had to learn from scratch and adapt the design to meet our project requirements.
Computer vision programming was another complex area that required extensive self-learning, testing, and iterations. Our team invested considerable effort in building from basic libraries to optimizing for more stable and accurate outcomes. This involved refining the Otsu binarization method for automatic image thresholding to better suit LentiMark detection and developing a custom angle calibration procedure tailored to our project’s specific needs.
Future impact
In computer-assisted surgical systems, precisely tracking objects like surgical instruments, anatomical structures, or mechanical devices is crucial. Instead of developing complex algorithms to identify and track each object individually, visual markers are attached directly. Once detected, the system infers the position and orientation of the associated object based on information from the marker. This method allows for precise control and accurate tracking while offering flexibility across different procedures and equipment with minimal adjustments.
Our project delves deeply into the design of the LentiMark visual marker. In the future, LentiMark could be used not only in surgical applications but also in robotics and beyond. LentiMarks presents a compelling alternative by combining enhanced accuracy with the compact size and durability needed for advanced applications. This positions LentiMarks as a superior option for a wide range of tasks requiring visual markers.
