Towards an integrated swarm intelligence framework for urban mobility: A systematic review and proposed theoretical model

In recent years, autonomous electric vehicles (A-EVs) have attracted the attention of academia and industry. In urban mobility, this topology requires consensus to control behaviours under swarm robotics. Although several model-based solutions have successfully enhanced accuracy and overcome some limitations, specific technological, methodological, and security issues remain. In this study, we systematically reviewed existing research related to swarm intelligence and multi-agent systems in urban mobility. Based on the obtained results, we propose a new directed acyclic graph-based multilayer architecture model. Furthermore, we propose a long short-term memory recurrent neural network model to make predictions. To validate the model, available data based on real traffic in Madrid (Spain) were considered.