Anti Fly-Tipping PNIPU

Main articles:

• Geographic information system
• Neuronets (neural networks)

2021: Announcement of "digital map" of unauthorized dumps

On February 9, 2021 it became known that researchers from the Perm Polytechnic University created "digital map" of unauthorized dumps. For the first time it is developed using neuronets. The application will help to find quickly illegal burials and to monitor them in dynamics – from origin before liquidation.

Determination of the area of a dump

According to scientists, in Russia and many other countries there is no full-fledged system of waste recycling yet. By estimates of Ministry of Natural Resources and Ecology, in the Russian Federation about 70 million tons of solid waste which accumulate on dumps are formed every year. Their area increases by 500 thousand hectares annually. By data Greenpeace, burn only less than 2% of all number of waste, and overwork – about 4%.

Example of work of neural network

For February, 2021 this problem is solved with the help of interactive Internet cards. On them any user can designate the place of an unauthorized dump. People can also collect statistics manually on maps of the land surface, using satellite services. But these methods are rather labor-consuming and require time and financial resources. Besides, yet there are no complex tools for identification and accounting of dumps, control of their status. Vadim Danelyan, one of developers, the undergraduate of "Calculus Mathematics, Mechanics and Biomechanics" department of the Perm Polytechnic University told

As it was explained, the service for monitoring of illegal dumps of Anti Fly-Tipping will allow to monitor quickly a situation in dynamics. Any user will be able to find burials at a stage of their origin, to monitor their growth and to learn about liquidation. "Digital map" will allow to collect important statistics on dumps in the specific area.

Neuronets at the same time classify satellite pictures of the land surface and find necessary objects. The artificial intelligence automatically detects a dump, determines its coordinates and estimates the area. Accuracy of a neuronet reaches 89%. The application will show how the status and the amount of burial changes eventually. Besides, it will be possible to define the owner of the territory by service, its cadastral number and automatically to make the claim. the developer added

Example of work of neural network

According to the young researcher, the application will be useful for state bodies and the ecological organizations. The service will help to estimate an ecological situation in the region and to control work of operators on the address with TKO. Also operators can use a product timely to reveal burials.

The web application which will help to trace statistics on dumps around the world will become an end product. For February, 2021 researchers develop and train neural network models to improve the accuracy of service.