Título: An Anomaly Detection Model Using GPS Data from Public Bus Systems in Large Cities
Palestrante: Mayurí Annerose de Morais (Doutoranda em Ciência da Computação – UFABC)
Data e local: Quarta, 22 de junho de 2022 às 16:00, Campus Santo André, Bloco A, sala S-204-0
Resumo: Providing efficient urban mobility is one of the main goals for using smart city technologies by public administrations. Disruptions caused by traffic incidents are an important cause of delays in public bus systems. Several Incident detection Systems (IDS) have been proposed for passenger vehicles, using multiple sources of information, such as camera images and vehicle sensors, including Global Positioning System (GPS). However, GPS data from city buses are much more scarce than from passenger vehicles, requiring different incident detection algorithms. This work proposes a bus system’s disruption detection model using bus GPS data. The model is composed of three components: a city-wide graph model maintaining the representation of the bus system as a graph; a bus position model, which collects real-time bus GPS data and maps their position into graph links; and the disruption detection model, where we combine the results of the previous models in a machine learning algorithm to check whether a link of the graph has an abnormal state.