The use of electric vehicles has gained momentum in recent times as they prove to be eco-friendly and energy-efficient. EVs offer a long-term solution to reduce the dependence on fossil fuels and greenhouse gas emission. Decreased air pollution due to the elimination of the exhaust pipe in electric cars promotes sustainable mobility. This in turn greatly reduces the negative impact of transportation on the quality of the atmosphere. However, uncoordinated charging of a large fleet of EVs poses serious challenges to the stability and security of the electric grid. Smart electric vehicle charging has recently gained significant attention in the research community due to the need to charge large number of electric vehicles economically. Not only should EV charging be economical, but also be energy-efficient and not tax the electric grid. Since the power demands of a building or residential area are not always constant throughout the day, the surplus power could be utilised by shifting time-flexible consumption such as EV charging to periods of lower demand of power. Ideally, EV charging load could be shifted to fill the overnight electricity demand valley while also considering the electricity tariff. In this thesis, a valley-filling scheduling algorithm is implemented that considers the Time-of-Use tariff to shift EV charging to off-peak hours and low tariff periods. The research also proposes a neural network model to predict future load based on weather attributes such as temperature and humidity. The simulation results demonstrate a good percentage of valley-filling achieved by the algorithm along with reduced tariffs.