The SUSU “Heart of the Urals” Advanced Engineering School announced the launch of a project called “Smart Eco-city Intelligent System: New Technologies for the Maintenance of Road and Transport Infrastructure”. The work is aimed at introducing modern digital technologies in the management of urban infrastructure, in particular, in the cleaning of road surfaces using municipal equipment.
The main objective is to study methods for assessing the quality of cleaning work and the efficient use of municipal vehicles using a high-tech hardware and software complex. The project for automated monitoring, selection of optimal operating modes of sweeping machines and assessment of cleaning quality is based on computer vision and deep learning methods using neural networks.
.jpg)
The AI-based system will control the quality of cleaning, allowing to accurately determine the level of road surface pollution and set the necessary parameters for the operation of the working bodies, such as the brush pressure level, the suction power of the vacuum cleaner and the zoning of the water supply. In addition, it will be able to identify large debris that can lead to equipment failure, which will increase the safety and reliability of cleaning work. The system will significantly reduce energy costs, which is especially important when using sweeping and cleaning machines with electromechanical transmission.
Currently, the team of scientists is collecting data to train the neural network. Specialists are identifying sources of information and image classes to prepare a dataset that will include various pollution objects recorded under different lighting conditions and on different types of road surfaces.
The system is ready to be integrated into unmanned vehicles, which opens up new horizons for automation and improving the quality of cleaning processes.
The project meets the goals and objectives of the “Formation of a Comfortable Urban Environment” national project.