Recycling and reuse of electronic waste is complex and cost -intensive – also because the disused electrical devices are difficult to disassemble and separate them into their components. In the Idear project, a German research team has now developed a method that is supposed to simplify this. For this, the old computers, televisions or washing machines are automated and dismantled with the help of artificial intelligence.
Manufacturing processes in the electronics industry are designed to produce particularly inexpensive – but not to re -use the materials and components as easily as possible. At the same time, the use cycles are getting shorter. This causes an increasingly greater amount of electronic waste. According to the Federal Statistical Office, around five million tons of outstanding electrical and electronic devices in the 27 EU countries hiked into the trash in the 27 EU countries. To a large extent, old devices – unless they are shredded – will also be recycled. To do this, however, they have to be dismantled manually, adjusted by pollutants, mechanically shredded and separated into different political groups by sorting processes. This manual disassembly is associated with high costs and is not very effective.
Cameras recognize components
The Fraunhofer Institute for Factory and Automation (IFF) in Magdeburg has therefore developed a system for automated and non -destructive disassembly processes of electronics in the Idear project. The approach combines measurement and robot technology with artificial intelligence and knowledge management. The name Idear stands for “intelligent disassembly of electronics for remanufacturing and recycling”. The method: After the electronic waste has been delivered and isolated, identification and finding is at the beginning of the process chain. AI-based 3D camera and optical sensor systems record labels with information about the manufacturer, product type and number. You also recognize the type and location of components, check geometries and surface, evaluate the condition of connecting elements such as screws and rivets and detection anomalies.
“Optical measurement technology helps to record labels and to sort different components such as screws,” explains José Saenz, group leader assistant, service and industrial robot at Fraunhofer IFF. Previously trained machine learning algorithms identify and then classify materials, plastics and components based on sensor and spectrald data in real time. According to Saenz, for example, the AI recognizes whether a screw is attached or rusted. All data is recorded in a digital dismantling twin, which is a kind of instance of the product and also informs whether a similar product has already been dismantled.
Disassembly from PC to the microwave
In the next step, Saenz and his team determine the disassembly sequences in software, which, among other things, define whether a complete or partial disassembly, i.e. only the recovery of high-quality components, should take place. The decomposition process starts on the basis of this information. The robot receives a number of instructions and processes to be processed, such as “Remove two screws on the left on the housing, open the housing”. The skills defined in the dismantling sequences include robotic actions such as screws, lifting, cutting, pulling out and localizing. In tests, according to Fraunhofer Iff, the demonstrator even succeeded in taking a mainboard out of a PC case-a very complex task that requires a high level of sensitivity.
The individual demonstrators for the sub -processes have already been built up. In the next step in the project, these should now be chained together. The goal is a demonstrator that integrates all technological developments and represents the chain of automated disassembly processes. The researchers are currently focusing on the automated disassembly of PCs. In the long term, however, other electronics devices should also be recycled more easily using the process. “In the Idear project, we strive for data-driven methodology so that from PCs to microwaves to white goods can be dismantled as different as possible products with low engineering effort and in real time,” says Saenz. Because recycling and re-manufacturing are a key so that producing companies can secure access to raw materials and to reduce environmental pollution from electronic scrap.
Source: Fraunhofer-Gesellschaft