UGV Digital Twin for garbage detection through AI
Agostino G. Bruzzone · 2022
The efforts to maintain low level of pollution in the environment need to meet sustainable and innovative solutions in order to be effective in the modern and complex world. Indeed, currently we are facing critical issues in terms of sustainability and new technologies are crucial to identify new solutions: recycling the abandoned wastes is a critical issue that needs to be considered and implemented beside their collection and processing. Due to these reasons, the purpose of this research is to develop an intelligent cleaning complex system based on UGV (Unmanned Ground Vehicle), which focus is to enhance the ability of strategic decisions making and achieve satisfactory trash detection in the wild for deployment. The system stands on AI (Artificial Intelligence) based on deep neural networks, which behaviour is simulated in a 3D Unity synthetic environment, through the usage of a digital twin, that simulates the desired solution of detecting and collecting garbage. The developed virtual environment chosen by the authors is a contaminated beach, in which an armed and autonomous robot manages to collect and recycle the wastes; the scenario is flexible to changes and further ones can be developed for training and testing of the intelligent system.