Design and Development of Fleet Management System

I designed and implemented a system that stores information about vehicles manufactured and delivered to customers in a database, enabling administrators to easily check each vehicle's status and warranty period. Customers can use the web-based operations system to monitor the real-time status of their vehicles in operation and remotely control them when necessary. The system is flexibly designed to be deployable as a cloud-based SaaS, on-premise, or even in air-gapped environments.

To support integration with existing systems, the architecture clearly separates the data layer from the vehicle control layer. The message structure was designed so that the system can infer vehicle status and event sequences solely from received data, regardless of whether all events are received.

Vehicle control functionality is implemented using an abstract unit called an "Operator", with each Operator responsible for a specific vehicle control protocol. All system components are required to implement a common set of APIs, enabling seamless portability across in-process modules and remote servers via a unified interface. As all related engineers shared the same API knowledge, components such as the Self-Driving Operator for autonomous driving features and the Multi-Agent Operator (VDA5050) for simultaneous control of multiple vehicles could be developed independently by different engineers and integrated into the system without friction. The centralized control system, along with clients that communicate directly with individual vehicles, is implemented from a single codebase.

The system supports OpenTelemetry, enabling transparent collection and analysis of internal metrics and external request flows.