The waste Management systems of urban areas are increasing in need of intelligent, scalable and sustainable solutions. The merger of Artificial Intelligence and the Internet of Things (AIoT) has been gaining more momentum in the prior decade in particular 2018–2025 in practical implementations of autonomous mobile robots (AMRs), smart bins with sensory capabilities, and edge cloud computing to end-to-end waste management. This systematic review, using PRISMA guidelines, synthesizes peer-reviewed literature from IEEE, Elsevier, Springer, Wiley, and MDPI to identify and screen studies and analyses perception (vision-based waste detection & multimodal sensing), cognition (on-device/edge inference, route planning, learning-based decision-making) & action (ROS-enabled autonomy, manipulation, fleet coordination). We connect key enablers (e.g., YOLOv4/Lightweight MobileNets) CNN/transformer detectors, SLAM and LiDAR/ultrasonic fusion, LoRa/NB-IoT/MQTT communications, and edge cloud orchestration to reported results in terms of classification accuracy, collection latency, travel distance reduction, and energy utilization. Recent prototypes and pilots suggest a range of steady advances made on proof-of-concept vision prototypes such as resource-aware edge inference, privacy-aware learning, and resource-aware telemetry at a city scale. However, there are still open gaps on robust operation in unstructured environments, interoperability among heterogeneous municipal platforms, energy sustainability (battery-solar hybrids, mission planning), and human-robot interaction (safety, acceptance, governance). We put forward a research agenda focusing on federated and continuous learning for AIoT; standardized data/ontology schemas; green autonomy (energy-aware planning and renewable integration); and ethics by design frameworks for public-space robotics. Through consolidation of developments and unresolved problems of 2018–2025, this review makes sense on a coherent taxonomy, comparative lessons and a forward roadmap to come up with scientifically rigorous, deployable and citizen friendly AIoT waste systems in smart cities.