Web and mobile app development for smart vehicle entrance & exit

Ong, Lip Wei (2023) Web and mobile app development for smart vehicle entrance & exit. Final Year Project, UTAR.

Preview

PDF Download (16Mb) | Preview

Abstract

The year 2023 is marked by a growing interest in the field of artificial intelligence (AI), which offers new ways to address problems and enhance efficiency. One area in need of a more efficient solution is the monitoring of car access records and management of car owners, which currently involves the manual processes by paper and pen. To address this, YOLOV4 model has been deployed to detect car license plates, and real-time web and mobile applications have been implemented to monitor car access logs and manage car owners. The applications offer filter and search functions, making the query process more manageable. Additionally, it includes role and permission functions to restrict access to certain features for security purposes. The web application is built using ExpressJS and Docker and deployed on two AWS cloud servers. HTTPS encryption and AWS Application Load Balancer are used to ensure security and reliability of web service. Furthermore, Github Actions is used to promote continuous integration and continuous deployment (CI/CD) flow. The React framework is utilized for the web application while React Native framework is used for mobile application development, both employing the Ant Design UI library to maintain consistency across screens. The Firebase cloud database is used for storing license plate pictures and providing real-time updates, while Algolia AI search service is employed for advance search functions. Finally, the project went through three testing processes, including unit testing, system usability testing and onsite testing. The project used unit testing to ensure that every function works properly, also 5 participants were invited to conduct the usability test of the system. The SUS score of the web application was 89%, and the SUS score of the mobile application was 89.5%, both of which were higher than the average SUS score of 75%, demonstrating good usability. Finally, the project also conducted onsite testing in a residential area to ensure that the system can run stably in the working environment.

Actions (login required)