Smart waste collection strategies have been developed to replace conventional fixed routes with dynamic systems that respond to the actual fill-level of waste bins. The variation in waste generation patterns, which is the main driver for the profit of smart systems, is exacerbated in the United Arab Emirates (UAE) due to a high expatriate ratio. This leads to significant changes in waste generation during breaks and seasonal occasions. The present study aimed to evaluate a geographic information system (GIS)-based smart collection system (SCS) compared to conventional practices in terms of time, pollution, and cost. Different scenarios were tested on a local residential district based on variable bin filling rates. The input data were obtained from a field survey on different types of households. A knowledge-based decision-making algorithm was developed to select the bins that require collection based on historical data. The simulation included a regular SCS scenario based on actual filling rates, as well as sub-scenarios to study the impact of reducing the waste generation rates. An operation cost reduction of 19% was achieved with SCS compared to the conventional scenario. Moreover, SCS outperformed the conventional system by lowering carbon-dioxide emissions by between 5 and 22% for various scenarios. The operation costs were non-linearly reduced with the incremental drops in waste generation. Furthermore, the smart system was validated using actual waste generation data of the study area, and it lowered collection trip times by 18 to 42% compared to the conventional service. The present study proposes an integrated SCS architecture, and explores critical considerations of smart systems.