One of the common issues in urban cities is traffic congestion especially at signalized intersection which may be caused by static and inefficient traffic light timings. The long waiting times at the traffic light led to significant amount of time loss, higher car emissions and higher fuel usage. During peak hours, the pre-set timing may be unfit to handle the higher traffic volumes than usual, which resulting in slow vehicle movement and increased travel time. Furthermore, the current technique used in setting the traffic light timing might not include spatial data or intelligent prediction tools that is capable to analyse the unpredictable characteristics of traffic in urban areas. This gap highlights the need for traffic light timings that is suitable for different traffic conditions throughout the day instead of constant fixed timing for all day. Hence, this study aims to derive optimal traffic light timing at junction by integrating Geographical Information System (GIS) and Artificial Neural Network (ANN) models. Unlike the current fixed-time signal control practice, the ANN was chosen to conduct predictive modelling that did not rely on human in making predictions. It is solely depending on data without relying on manual assumptions or fixed timing. The study used historical and current traffic volume data as well as the existing signal timing parameters to develop ANN models to predict optimal green time allocations based on different traffic demand patterns at the intersections.