More information
Description
This project, led by Polymore SRL, a dynamic IT startup, tackles, in collaboration with the EEA and Norway Grant the challenge of improving waste collection processes in Romania. By leveraging cutting-edge technology, the project seeks to provide direct and customized feedback to waste generators, be it individuals or companies. This feedback mechanism aims to enhance the quality of recyclable waste sorting, increase transparency in waste collection actions, and align waste collection strategies with local needs. Ultimately, the project endeavors to boost recycling efforts, reduce waste contamination, and contribute to a more efficient waste management system in Romania.
The project’s main aim is to develop an IT solution which will unlock waste collecting processes, by addressing direct feedback to waste generators. By completing this project, PolyMore is closer to the goal of increasing waste recycling rates in Romania. This will lead to improved competitiveness for Polymore SRL, as well as improving the environmental focus in Romania.
To achieve this targets, the project takes on a series of key activities, including:
- Project Management
- Project Promotion
- Participation at waste management events and conferences
- Development of Innovative IT Solution
Summary of project results
The project was designed to address several critical challenges in the waste management sector. One of the primary issues was the lack of direct, customized, and immediate feedback for waste generators, both individuals and companies, regarding the quality of their recyclable waste sorting. This absence of feedback led to a lack of transparency in the actions of waste collectors and a consequent disinterest in the recycling system. As a result, the quality of collected waste batches was often low, contributing to poor results in waste sorting activities at processing centers in Romania.
Additionally, the project aimed to tackle the lack of adaptation to local waste collecting strategies. Each local area typically has specific bins and waste collection strategies, but there was no direct feedback mechanism for each waste generator. This made it difficult to ensure consistent and correct waste sorting practices.
The project proposed the use of high-end technology to enhance waste collection processes. By engaging with waste collecting companies that operate in the B2B sector, particularly those involved in industrial waste collection, the project sought to improve the overall efficiency and effectiveness of waste management practices. This technological approach was intended to provide immediate, customized feedback to waste generators, fostering greater transparency, accountability, and interest in the recycling system.
The Promoter has developed 1 GPS-based neural network that calculates the level of inefficiency in waste collection activities. This outputs a data-based model for taking actions and improving the operational strategy for the company''s clients. In addition, Polymore has developed 3 neural networks for tracking the important feedback elements regarding the correct sorting of waste:
1. Detection of waste container being unloaded onto the waste truck: with an accuracy of 97% and precision of 82%, the model is deployed to the company''s local servers built internally.
2. Classification of waste materials present in the waste truck after container unloading: with an accuracy of 73% and precision of 72%, the system classifies the materials in the 30 frames per second videos, after trigger from neural network 1.
3. Neural network for classification of foreign objects - large contaminants: with an accuracy of 97% and precision of 93%, our model detects large
contaminants and alerts waste collecting companies of irregularities.