How Electrocycling GmbH manages to identify and sort out batteries and old appliances containing batteries as well as harmful and contaminating materials when processing electronic waste.
The WeSort.Al technology analyzes the incoming e-waste, increases the sorting purity for the subsequent treatment processes and thus prevents potential fires caused by overlooked lithium batteries.
With a processing rate of up to 80,000 tons per year, Electrocycling GmbH is one of the largest recycling companies for waste electrical and electronic equipment in Europe. For more than 30 years now, Electrocycling, based in the district of Goslar, has been recycling old electrical appliances and producing high-quality secondary raw materials.
"The waste delivered often contains batteries or materials that are harmful to the environment. If we overlook these during manual pre-sorting, this can result in contamination or fires in the later processes."
-Kai Kramer
â– PROBLEM
Kai Kramer
Leiter Qualität
â– SOLUTION: FIRST STEP
Before the recovery and recycling steps, the old appliances delivered from private households and municipal collections must be sorted. This sorting is carried out manually on a sorting belt by up to 8 employees per shift. Components and appliances containing harmful substances such as batteries, fluorescent tubes, smoke detectors, capacitors, mercury, etc. are sorted out. In addition, contaminants and waste consisting of commercial waste, wood, textiles, glass or ceramics are removed.
Electrocycling GmbH uses the WeSort.Al analysis module for optimum identification of the materials and analysis of the material flows. The module was installed at the very beginning of the manual pre-sorting process and uses AI algorithms to recognize a wide range of material classes and display these findings in real time on a dashboard. This makes it possible to continuously monitor the material flow through the system.
The core component of the system is the AI sensor unit, which consists of a digital camera system. By using deep learning for object recognition, materials and objects can be identified with high precision. The sensor unit serves as a central interface for the analysis, sorting, control, optimization and monitoring of the sorting process. It is characterized by cost-efficient hardware, simple installation and its effectiveness in various applications.
The underlying AI software of WeSort.AI, based on deep learning for object recognition, supports the creation of dashboards for uncomplicated real-time monitoring of material flows. Customizable alerts provide immediate information about malfunctions, potential hazards and unusual events.
Devices and materials that are reliably identified by the system include Batteries and battery-containing devices such as tools, small household appliances, smoke detectors, hygiene devices, transponders, measuring devices, heat meters, etc. as well as metals, plastics, textiles, wood and glass.
​
â– INSIGHTS: FROM THE IMPLEMENTED SYSTEM AT ELECTROCYCLING GMBH
The system offers a real, preventive solution for fire prevention and also addresses a variety of other financial and operational challenges.
The introduction of the system in sorting plants reduces minor fire damage, which costs an average of 200,000 euros per year and plant. It also protects against the potential total loss of entire sorting plants, which can result in damages of up to 30 million euros.
"The AI system from WeSort.Al can also reliably detect hidden or built-in hazardous components such as batteries. It helps us to better understand what is running through our system and increases our system safety."
-Kai Kramer
Kai Kramer
Leiter Qualität
â– SOLUTION: SECOND STEP
In addition, the material flow is analyzed by an analysis module at the end of the sorting line. This module serves as an efficient warning system that precisely identifies which potential contaminants and impurities may have been overlooked in the first sorting phase, i.e. manual pre-sorting. If such substances are detected, the system immediately triggers an alarm to enable subsequent separation. The system also provides the shift supervisor with continuous updates on the quality of the sorting. Particularly critical materials such as batteries, which pose a high fire risk, can be reliably identified, significantly increasing safety.
Furthermore, the aim is to find an automated sorting solution for e-waste in order to ensure effective re-sorting of the identified objects. The modular system from WeSort.AI offers an excellent opportunity to expand the analysis module with a highly effective sorting bar. This pneumatically operated air pressure nozzle bar can be easily and cost-effectively integrated into existing systems and can be flexibly installed at the outfeed of conveyor belts at speeds of up to 4 m/s.
76 % increased safety
50% purer sorting
Annual EBIT increase of over 600 thousand euros
"The aim of further joint collaboration is to fully automate manual pre-sorting."
-Nathanael Laier
Nathanael Laier
CEO WeSort.AI GmbH
