Installing wireless sensors on a network of waste containers, that monitor at regular intervals their waste fill levels and transmit the data to the cloud of a waste management operator over the internet, describes the Internet of Things (IoT). With this constant stream of data, the dynamic organization of waste collection schedules is enabled as containers are visited for collection only when it is necessary, which consequently leads to demand-responsive services. Domain experts attest that operating a demand-responsive service brings financial and environmental gains to a waste collection service, but it simultaneously introduces complete variability in the system which is undesirable in real-life operations. To solve this problem and consequently optimize the waste collection service’s performance, domain experts stress the need for a balanced trade-off between dispatch consistency and flexibility. This means, being able to exploit to the highest degree possible the benefits of demand-responsive operations, while also maintaining a certain level of dispatch consistency when demand varies from day to day. This research focused on developing a solution approach to solve the IoT-based waste collection problem, which is derived from the knowledge and requirements of the domain. The use of the knowledge of the domain is significant as it ensures that the model is tailored and applicable to a real-life IoT-based waste collection service. The overarching objective of the approach is to maintain dispatch consistency and flexibility when the containers’ location and demands vary from day to day, as well as to attain an economically and environmentally enhanced waste collection performance.