Online Monitoring of Buses Information Using KNN, ATR and DMC Algorithms

Abstract

In developing countries such as Brazil it is still a challenge to know what time the busses will arrive in certain busstop. This makes millions of people wait for long periods with no certainty that the bus is near or far away, causing urban mobiblity challenges that affects everyday millions of public transport users. One way out is to search solutions based on software and artificial intelligence that can help users during their everyday-jorney. However, most of the options still need to be approved by private companies or government. In order to propose a new alternative that is non-dependent of companies or government, this article proposes the online monitoring of bus routes in a collaborative way, i.e., among users, by using adapted K-Nearest Neighbors algorithm and two new algorithms called Activation of the Regressive Timer and Deactivating Collaborative Mode. The algorithms were implemented in one application named “Olha o Ônibus” and seven tests were performed to prove the efficiency of the proposal. The results indicate that the proposed algorithms can guarantee the reliability of the information shared among users.