Development of API Service to Store and Utilize Video Analytics Data from Edge

Abstract

In every train arrival at the station, the recording of the train arrival is still done manually by the officer to the Pranopka system. The manual operation has the possibility of errors due to negligence, non-standard time referenced, and the absence of parameters to measure its correctness. Therefore, it is necessary to build an automatic recording system in order to eliminate the burden of repetitive tasks to humans. We designed an automatic recording system using a camera. This paper will explain in detail about the architecture development of the API used for data transmission in managing data that is in the database in a centralized data center. In its development, various literature analyzes were carried out to determine the most appropriate technology to be used in the system. From the development process that has been carried out, it has been proven by non-functional testing that the use of MySQL, MongoDB, Fastify, JWT, PM2, and Nginx is suitable for developing an automatic train number recording system. The throughput can reach 1,000 transactions per second with APDEX 0.983.