Industrial Automation and Processes

Project AI-47: AI Accelerator Architectures Incorporating Built-In Self-Test, Self-Diagnose, and Self Repair

Project Client: UBC ECE SoC Lab

Project Description: Deep neural networks (DNN) accelerators are the latest trend in advancing computing devices and are used in domains where errorless decision-making and operation is paramount, such as medical technologies and manufacturing industries. Systolic arrays are fundamental building blocks in many such accelerator architectures, and thus require robust fault tolerance mechanisms to ensure flawless operation under every usage scenario.

To address this, our project implements a systolic array design that is capable of self-testing, self-diagnosing, and self-repairing permanent faults without imposing significant repercussions on area, power consumption, or overall performance of the entire systolic array.

Project HA-83: Sizing of Temporary Personal Protective Grounding Cables

Project Client: BC Hydro

Project Description: In British Columbia, BC Hydro is the electric utility provider for 95% of the population and serves over five million people in the province. Electricity is delivered to homes and businesses via transmission lines spanning thousands of kilometers. Operating and maintaining this intricate network of transmission lines is a difficult task. When the line needs to be maintained, there is a coordinated shut down to reduce the disturbances of power loss. Once this is determined, the lines are de-energized at the connecting substations, then the line workers can proceed to perform their maintenance tasks. When working on these lines, the line can accidentally be re-energized, causing a voltage potential to induce current into the line. To protect the workers, a temporary personal protective grounding wire is used to ensure the voltage is at a safe working level and to provide a path for current to flow. This grounding wire needs to be sized correctly to withstand the current for a period without melting until the protection operation is initiated. Traditionally, the fault current and grounding wire is calculated manually, which is very tedious, inefficient and prone to human error. 

Our solution is an Excel Program and IOS application that automates this cable calculation process. Both the excel program and application are integrated with BC Hydro’s existing database of lines and stations. These databases provide information, such as impedances and lengths, so that the fault current can be calculated at any specified location. The programs are built to work as simply as possible for the user. The user will simply select the line that they are working on, input the required and optional fields and the program will automatically run the calculations,  outputting the required grounding cable size. Engineers in the office can use the Excel program to calculate the cable size and validate the calculations that are being performed in the program. The application is a portable tool for workers in the field to reconfirm the appropriate grounding cable size needed. The grounding cable is physically demanding to install due to its length and weight, by using the programs we have developed, the workers can determine the minimum cable size needed at their specified location. This would reduce the redundancy of using an oversized grounding cable while maintaining the safety of the worker.

Project HA-84: Medium Voltage Underground Cable Circuits Ampacity Tool

Project Client: BC Hydro

Project Description: The client for this project is BC Hydro. They requested an easy-to-use Excel tool to perform ampacity calculations for medium voltage underground power cables (up to 25kV). These calculations determine how much current can be passed through cables while staying within safe operating temperatures.

These calculations consist of almost one hundred variables and formulas which can change depending on many environmental factors. Historically, engineers have performed these calculations by hand or used expensive software to comply with industry standards. We have developed our program to simplify this process by requiring approximately twenty inputs. These inputs are given to our calculation engine to complete the study almost immediately. This provides time saving benefits to the engineers at BC Hydro.

Project HA-85: Frozen Soil’s Impact on the Thermal Resistivity’s Used in Medium Voltage Cable Circuit’s Ampacity Calculations

Project Client: BC Hydro

Project Description: Purpose: Investigating the effects of temperature change upon soil resistivity. This project is designed to highlight the effects of extreme temperature changes (such as freezing) on top soil layers that are used inside substations, to our client, BC Hydro. This evaluation of soil resistivity is crucial for safety purposes of substations, for the following reasons: 1) Provide means to carry carry current into earth under normal and fault conditions without exceeding operating conditions or equipment limits 2) To reduce the risk of a person in the vicinity of grounded facilities being exposed to the danger of critical electric shock.

Major Design Contribution: Our team designed a lab apparatus that can contain soil samples from substations in a watertight setting at large volumes. This enables any future study by BC Hydro to be done in lab conditions, instead of in field. Samples can be brought from all substations to a singular lab and be tested in different temperature conditions for further evaluations. 

Louise Baldry 778-835-4412; Ali Ghahremani 778-834-2158; Ali Salehi Najafabadi 778-512-593; Omar Abdelmouti 778-386-5432; Shawn Wali 778-707-9416

Project JY-71: 3-Axis Motion Control Testing System

Project Client: NovaSense Technology Ltd.

Project Description: Pressure injuries are a common occurrence for bedridden and mobility-challenged individuals, stemming from a lack of blood flow. These injuries are easily avoided by shifting one’s position when discomfort occurs, however in many cases, patients may not be able to communicate or perform this action on their own, and the responsibility falls on nurses and healthcare professionals to both detect and adjust for pressure ulcers. 

Our client, Novasense Technology Ltd. has developed a flexible pressure sensor technology known as the Smart Sheet, intended to be placed in the beds and chairs of mobility-challenged individuals to help detect and prevent pressure ulcers. In a desire to further test the technology, our team was tasked with the creation of what has been named the Smart Sheet Characterization Unit (SSCU), a device capable of applying normal and shear forces to the Smart Sheet, as well as recording data which will be invaluable in further development of the Smart Sheet.

The SSCU is a 3-axis cartesian robot that features a bed capable of supporting the Smart Sheet which can move in both the X and Y planes, as well as a pressure applicator mounted to the Z axis, which can apply a vertical force to the bed. Allowing the bed to move in two axes, allows a shear force to be applied in any direction in the XY plane, as the bed can move while a force is applied from the Z axis. The pressure applicator is capable of accepting user-customizable indentures, allowing for testing of different contact surfaces against the Smart Sheet.

The hardware system connects to a software interface running on a PC, allowing the user to create custom test profiles, including multiple options for force application and bed displacement. The software is also responsible for recording data from the SSCU and displaying it to the user, such that a test may be monitored in real-time, and that test results may be further analyzed afterward.

These features will allow precise normal and shear forces to be applied to the Smart Sheet, such that its performance may be analyzed and characterized. This data will accelerate the development of the technology, and help step towards the deployment of the Smart Sheet.

Project JY-72: 3-Axis Motion Control Fabrication System

Project Client: NovaSense Technology Ltd.

Project Description: Our client, NovaSense Technologies Ltd. is a Vancouver-based medical device company dedicated to devising innovative sensing solutions for complex healthcare problems, with a primary focus on pressure injuries. Through the integration of intelligent sensing technologies and personalized user solutions, NovaSense is committed to enhancing the quality of life for individuals affected by and susceptible to this persistent health concern. Their flagship product, the Smart Sheet, is a soft, flexible, and stretchable sensor array designed for continuous pressure monitoring that can be integrated with wheelchair support cushions and/or bed mattresses. Coupled with their prevention algorithm, an alert system provides real-time insights on high-risk pressure injury locations and time to healthcare professionals, caretakers and users, enabling immediate preventative action to prevent the formation of this complex chronic wound.  

The current fabrication process of the stretchable electrodes involves a rather labour-intensive and time-consuming approach that includes stencil masks and spread-coating techniques. This fabrication method is susceptible to human errors, resulting in quality variations affecting overall performance. Therefore, the main goal of this capstone project is to design a fully automated three-axis motion control fabrication system that is capable of streamlining the Smart Sheet manufacturing process, which the team has named the Capacitive Sensor Printing System (CSPS).  

The CSPS is a fully custom-made device designed to make the fabrication process of the Smart Sheet as efficient and consistent as possible. For its main frame, it utilizes aluminum profiles which allow for a robust, highly modular and expandible platform. For printing efficacy, it uses a ball screw-driven ink extrusion system and linear actuators that allow for precise and consistent prints. The extrusion system has been designed to be incredibly user-friendly, as all the user needs to do is load a standard syringe with the liquid ink and simply click fit into the pumping base. To satisfy the client’s future needs, the system can also print ink in a wide variety of viscosities (ranging from acrylic paint to honey) in neat straight parallel lines even on slightly skewed surfaces; a task that is very challenging when working with liquids. The system also features proprietary software and a user interface. The software allows the user to change printing parameters, visualize these changes in a preview window and save these settings into a g-code file which is exported into an SD card. Adjustable parameters include sheet size, distance between parallel lines, start position, dispensing rate and many others.  The user interface incorporates the SD card and allows the user to select the file and begin printing. Lastly, the system also has a built-in curing system which can rapidly cure ink with lower viscosities by applying direct heat after printing.    

Project PB-35: Automated Test Framework Migration

Project Description: Team PB-35’s Automated Test Framework Migration Tool helps software development teams to streamline and enhance their testing procedures within the software development lifecycle, specifically targeting the challenge of migrating outdated testing codebases to more modern, efficient frameworks. This project addresses a critical issue faced by teams where maintaining a large, antiquated test automation codebase becomes an engineering burden. By automating the conversion of FitNesse SLIM tests into Java unit tests, the tool offers improvement in testing efficiency, cost savings, and resource reallocation, allowing teams to focus on more pressing tasks such as improving test cases and developing new features.

The major technical challenge and design contribution of this project lie in creating a robust command-line utility that can parse and transform the domain-specific language of FitNesse SLIM into Java, while ensuring the functional equivalence of tests. This involves sophisticated syntax analysis, transformation, and code generation techniques to handle various complexities associated with the translation process including the intricacies of the FitNesse language. Our approach ensures the tool is flexible, extendable, and capable of producing detailed conversion reports to assist in debugging and validation efforts.

Elise Hlady:; Jessica He:; Jack Le:; Josh Wong:; Jefferson Zhai:

Project PL-12: Orbital – Passive Antenna Surveillance System (O-PASS)

Project Client: MDA Systems Inc.

Project Description: The Orbital – Passive Antenna Surveillance System (O-PASS) aims to explore the feasibility of a ground-based passive RF system for monitoring geostationary satellites to improve MDA Systems space domain awareness (SDA). 

O-PASS is equipped with an antenna dish, optical telescope, tracking mechanisms, and deep-cycle battery, in a compact and portable subsystem design. This design allows operators to transport the system in a passenger vehicle and assemble on site in under five minutes. After O-PASS has been assembled, the operator calibrates the mount using the optical system and initiates the data collection process.

To collect data, O-PASS converts the downlink carrier signal of satellites in orbit into I/Q signals using its RF subsystem. This RF subsystem comprises a C-Band antenna, a low noise block down-converter (LNB), and a software defined radio receiver (SDR). The SDR software then performs a series of digital signal processing functions, including filtering, windowing, and FFT, to identify the center frequency, bandwidth, and amplitude of the carrier signal. This information is recorded on an internal storage drive and displayed to the operator for further refinement.

O-PASS is also equipped with a heated deep-cycle battery to enable operation in off-grid environments in temperatures from –20°C to 60°C, while being waterproof up to IP33. With efficient thermal management and energy dense lithium battery technology, O-PASS ensures continuous operation for extended durations, making it a promising tool for satellite monitoring and analysis.

Andrew Bishop:; Manik Dialani:; Qingzhe Liu:; Quentin Funk-Foster:; Sheng Jiang: