Proper waste management in urban areas is very important for the development of smart cities, hence the need to adopt effective, efficient, and sustainable technologies. Some common problems associated with ordinary waste management services include but not limited to; improper collection frequency, spillover events, and high resource utilization. In the following paper, we present an IoT framework for waste management that utilizes sensors and machine learning approaches to improve supervision and collection trips while at the same time increasing productivity. The system under consideration provides for the use of sensors to constantly measure the waste and environment parameters as well as forwarding the obtained data to a special server for further identification of specifics. For advancing predictions and route planning, artificial intelligent models are used and collection schedules and routes are transformed according to new data received. A detailed analysis of the implementation of the proposed system is presented through simulation experiments as well as real-life applications, proving significant increases in the efficiency of waste collection, decreased costs, and improvements in environmental impact. From the findings made from the study, it is understood that IoT and machine learning innovations have the possibility to greatly revolutionize the nature of urban waste management and, therefore, advance smart cities.