Mohamed S.I.University for Modern Sciences and Arts MSAEgypt2020-01-092020-01-09201922778616https://t.ly/b29YGScopusEgypt�s current bus system is large enough to satisfya significant portion of the populations demand, yet fails to do sodue t o mismanaged resources. Static lines and unclear schedulescreate a confusing and unappealing user experience which pushesmore of the population to cars for their transportation needs. Thisclearly leads to more congested streets which result in a net lossof productivity as well as an inc rease in stress, unnecessary fuelconsumption, and harmful emissions. An intelligent bus solutionis multi-faceted. It consists of (1) connected buses which arecapable of providing their geo-location data, feedback aboutdriving behavior, and health data to detect fail ures before theyoccur; (2) cashless payment through RFID cards to ensure muchtighter control over pricing; (3) a processing server or cloud, inwhich all of the incoming data would be handled; (4) knowledgesystems which dynamically optimize bus schedules and routesthrough learning algorithms; (5) and a mobile application tocapture demand and inform passengers of bus arrival times.The main functions and algorithms of the proposed systemare achieved based on machine learning algorithms and webtechnologies, whilst the hardware component is implementedbased on System-on-Chip technology with custom hardware tointerface with the vehicle. This paper will focus onthe software component of the proposed solution. It is shown that byapplying the proposed system to a previously static bus systemthat fuel consumption, maintenance costs, and carbon emissionscan be reduced by 10-20% while overall passenger satisfactionis increased. �2019.EnglishNeural networksParticle swarm optimizationUrban transportationVehicle routing problemWeb applicationArticle