Sam started his academic studies by earning a Bachelor's degree in Mechanical and Electrical Engineering from the University of Bristol. Subsequently, he pursued a Master of Science in Automotive Engineering with Electric Propulsion at the University of Bath. Sam's enthusiasm for the automotive industry grew during this course, and it was during the culmination of his academic journey, while working on his thesis project, that he discovered the perfect intersection of his mechanical and electrical knowledge. This newfound interest led him to the realm of computer vision in automotive applications.
For his Master's dissertation, Sam delved into the realm of autonomous risk assessment using computer vision applications, with the aim of enhancing safety in both driver-assisted and autonomous systems. His interest in advancing this field prompted him to join AAPS, where he aspires to contribute to the integration of autonomous systems into our daily lives.
Computer vision has the potential play a crucial role in almost all autonomous driving and traffic management systems of the future. This technology involves the use of cameras and image processing algorithms to interpret and understand the surrounding environment, in the context of automation, allowing the more efficient and accurate management of traffic, automatic crash detection systems, parking management and autonomous driving applications.
In this research project, the primary objective is to enhance the capabilities of computer systems in understanding and predicting the behaviour of vehicles on the road, with the ultimate goal of improving road safety and efficiency. The project focuses on two critical aspects: vehicle perception and object tracking. This is the observation of vehicles and important information about them, colour, vehicle type, shape, size etc, and the correlation of these properties across video frames in order to follow the vehicle over time.
The aim is to develop advanced computer algorithms capable of accurately identifying key attributes of vehicles, such as their movements and intentions, in real-time. This understanding of vehicle behaviour will contribute to safer driving scenarios. Additionally, the project seeks to improve existing object tracking algorithms by incorporating contextual information, like lane detection, to enhance trajectory prediction and situational awareness. The involvement on contextual clues specific to automotive situations should allow the algorithms to provide a more robust and reliable result that more generic algorithms.
Sam's project will be completed using a mix of analytical and machine learning algorithms. Where the two different approaches will be compared against each over for speed accuracy and ease of use. In an attempt to find a solution that can both provide usable results in a real-world scenario but also run on systems capable of being deployed.