本文へ

B1-03 Building a data-driven advanced sorting system

Principal Investigator

ISEKI Yasuto (Mitsubishi Electric Corporation)

Background and purpose

Build a data-driven “Smart Electrostatic Separation” system for stable supply of high-quality recycled plastic.

Schematic diagram of a data-driven advanced sorting system that uses static electricity to sort plastics.

Fig.1 Principle of electrostatic separation
(Example of our home appliance plastic recycling)

Research and Development Overview

Overview diagram summarizing the interrelationships between 1. Development of electrostatic separation equipment, 2. Development of sensors and a condition control system and 3. Development of information interaction interface.

Fig.2 Data-driven electrostatic separation equipment

Progress and Achievements

We have developed a verification equipment of electrostatic separation that can automatically sort mixed plastics according to changes in composition ratio. The hyperspectral camera installed in the verification equipment can identify the type of plastic with 99% accuracy and visualize the composition ratio of plastics before and after sorting in real time.

In addition, originally developed specific charge distribution evaluation system can sense the specific charge of plastics with high accuracy. By having AI learn the composition ratio of plastics before and after sorting and the specific charge distribution, it is possible to automatically calculate sorting conditions with higher purity.