Projects

Our team works closely with industrial partners. The team applies their expertise to address companies' specific technical problems.

  • Development of Lab on a Chip for Rapid and Accurate Point of Care Detection of COVID-19

The goal of this partnership is to develop a cost-effective and accurate point-of-care testing method by combining the relevant skills of the applicants and the industry partner. To achieve a rapid and on-site portable diagnosis of COVID-19, we will develop a highly sensitive near-patient diagnostic device based on thin-film transistors (TFTs) electrical sensing. This device will detect COVID-19 using minimally processed patient samples. The AMS lab is part of a multidisciplinary research team that is composed of researchers from the Mechanical Engineering, and Chemistry and Bio-Chemistry departments.

This research is sponsored by NSERC Alliance Grant and is in partnership with APAG Elektronik Corp, a national leader in the manufacturing of electronics and lighting components. (Ongoing)

  • Automated and Connected Electric Vehicle Integration- Detection of Hardware Trojan by Using Machine Learning

Electronic systems have advanced over the past few decades to the point that almost all of our daily activities depend on them. In doing so, we trust the Integrated Circuits (IC) within the electronic devices to perform their required operation.

However, due to current manufacturing trends, these ICs are outsourced to third parties, and therefore the issue of trusting the integrity of the IC has turned into a central challenge for the security and safety of systems. Without trusting these ICs, the systems that rely on them are open to attacks by a malicious adversary. The hardware and internal structure of the ICs can be modified, without the knowledge of the designer. The malicious, undesired, intentional modification of an electronic circuit or design, which results in the incorrect operation of the electronic device, is called “Hardware Trojan.” This research proposal investigates the development of a machine learning method that relies on the side-channel signature of the IC in order to detect the Hardware Trojan.

The research is funded by the MITACS and is in collaboration with the Canadian Urban Transit Research & Innovation Consortium (CUTRIC). (Ongoing)

  • Investigation of Hardware Security Issues in Using Electric Vehicle Fleets with Battery Exchange Infrastructure

Sustainable public transfer systems are getting a lot of attention, and many organizations are transforming their vehicle fleet into systems that rely on alternative fuels. Among these, the Electric Vehicle (EV) is the most popular choice.

The environmental, geopolitical, and financial advantages of EV vehicles are well studied and addressed in many research publications. However, the security of these systems is not given the full attention that it requires and will be the subject of this research.

The research is funded by the MITACS and is in collaboration with the Canadian Urban Transit Research & Innovation Consortium (CUTRIC). (Ongoing)

  • Windsor-Essex (WE) Diversify - Advance Auto mobility Technologies (Connected, Autonomous, Shared & Electric)

The integration of physical and hardware security aspects into automotive specifications is a growing issue. Physical security is an issue at the implementation of electronic for Vehicle Communication.

A dedicated hardware security module is required for the storage of private information and execution of cryptographic operations in a physical format. It should be noted that a comprehensive solution for all types of possible attacks and scenarios is a challenge that cannot be addressed in its entirety in this proposal. This research is supported by the FedDev Ontario and is in collaboration with the WindsorEssex Economic Development Corporation. (Ongoing)

  • Hardware Implementation of Bio-Inspired Systems for Real-Time Signal Processing

Many applications such as image processing, object recognition, probabilistic inference, or speech recognition that require a significant amount of high-dimensional data processing experience bottlenecks due to the separation of memory and processing elements. In these cases to be able to solve the problem efficiently, other systems that are based on conceptually robust yet straightforward and highly parallel systems such as neural networks and Deep Neural Networks can be employed.

Since digital arithmetic and implementation style has been established in many electronic systems, the central attempts in implementing the neural networks have been to use this format towards parallel processing. However, a massive number of interconnections required by the networks has caused issues with the digital implementations and has forced the signal processing towards time-multiplexed or serialized signals, which is in contrast with the original idea of implementing simple and highly interconnected neural networks. Therefore, it can be argued that the analog and mixed-signal systems can offer a natural way of implementing these systems. With these methods of implementation, the memory and processing elements can be localized, and variables can be represented by analog or multiple-valued digits instead of being encoded digitally.

This research is funded by NSERC, through the Discovery Grant Program. (Ongoing)

  • Development of the Predictive Digital Controller Based on the Optimization of In-Memory Operations in the FPGA

In recent years, with the advances made in the Field Programmable Gate Array (FPGA) powerful digital controller can be developed for the Power Factor Correction in power electronics circuits. However, these available solutions are based on merely programming the board and integrating it with the rest of the system. At this time, little attention has been given to possible improvements in increasing the efficiency of FPGA board operations, its acceleration, and reducing redundancy.

The approach here is to create a predictive digital controller and its feedback loop, optimizing the operations required in a way that does not result in slowing down the rest of the system. The feedback loop is used for compensating for possible errors when the system is operating in conditions that are not nominal. It is required for the output to adjust to changes at its load rapidly and efficiently.

This project is funded by both NSERC and Liburdi Automation Inc. (Ongoing)

  • Development of an Small Form Factor Camera for Welding Applications

This is aims to develop an imaging system useful for welding applications that can withstand harsh lighting conditions. Mainly the camera should be able to capture images at the start of the operation when arc light starts. General-purpose cameras are showing only white screens and cannot capture or record any images for few seconds. The sensor requires to have high dynamic range, around 140 dB, and requires to have accessible memory register. The digital registers are used to modify the algorithm using the DSP boards. This project was funded by NSERC, and in collaboration with Liburdi Automation Inc. (Completed)

  • Inspection of Highly Reflective Surfaces Using Machine Learning

The objective of this research project is to investigate and propose an automated inspection system that can be used for detection of the faulty automotive parts. Since the current method of inspection is human-based, it is not free of errors. Moreover, the current inspection method lacks consistency; the same part can be inspected by two operators with different results. The faults are mostly on the surface, are thin, and do not run deep into the base of the part. They are highly irregular in shape, size, and dimensions, and do not have the same pattern. Moreover, the more common missed defects are hard to catch by inspectors, and sometimes not easy to see. This project was funded by NSERC, OCE, and Windsor Mold Group. (Completed)

  • Secure Car to Car Communication

Over the past few years, there has been significant development in vehicular communications to provide active safety applications such as collision warning, collision avoidance, emergency vehicle signaling, roadside conditions, warning on environmental hazards, and traffic and road conditions. Any security weakness in the vehicle communication system has a direct impact on other vehicles and as well as on its surrounding environment. The purpose of this project is to develop custom-designed hardware that is secure against side‐channel device attacks while ensuring the efficiency of the design. This project is funded by FedDev Ontario, and in Collaboration with the Cross Border Institute. (Completed)

  • Online Airbag Fabric Quality Control

The aim was to investigate the development of an online quality control solution for airbag fabric manufacturing lines. We found a visual indicator that could be used to separate the acceptable and failed samples while the fabric was on the loom. A vision-based processing algorithm was developed which was used to alarm the operator of non-ideal conditions. The tests showed that the method is highly reliable and accurate, real-time, and non-destructive. Support for this project was funded by NSERC and in collaboration with Autoliv Inc. (Completed)

  • An Integrated Smart Active Safety System for Vehicles

The objective of the proposed research is to integrate sensors, advanced microelectronic signal processing, FPGA, vehicle networking, fault- tolerant integrated chassis control (FTICC), safety-related telematics, and data fusion algorithms to realize a dedicated micro-controller based Integrated Smart Active Safety System (ISASS) for vehicles. This project was funded by AUTO21. (Completed)