GPS-based Arrival Time Prediction for Bus Schedules: Enhancing Commuter Experience

The use of buses for transportation is significant in contemporary life and is becoming more and more widespread. The usage of private automobiles, fuel consumption, and traffic congestion are reduced if the buses arrive on time. The capability to predict real-time arrival times is highly beneficial for both passengers and transportation authorities. It cuts down the amount of waiting time they have to endure while travelling and gives them the gratification of being aware of the bus timetable. This demands a model that can inform the passengers about the arrival time of the vehicles, so that they can make their transit plans. With the development of this model, more people could be attracted to the public transportation which then reduces the congestion of roads due to excess private vehicles. In recent times, the proliferation of wireless communication technologies and the widespread use of Global Positioning System (GPS) have resulted in the generation of vast amounts of vehicle trajectory data. This data opens up new possibilities for accurately predicting real-time vehicle arrival times. In this paper, we propose a modern public transportation with IoT enabled system to give an accurate prediction of arrival time and also the location of the transport or vehicle to the particular bus stops. In this context, the Long Short Term Memory (LSTM) neural network is being employed. It is trained using various traffic parameters and environmental conditions to enhance its prediction capabilities for real-time vehicle arrival times. Parameters that are considered in the proposed system include distance, waiting time at stops, traffic density and road junctions.