AI Plastic Sorting Using NIR Spectroscopy and Blockchain Markers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The recycling of plastic waste is hindered by the need for effective sorting of various polymer types, grades, blends, and additives, which is not efficiently addressed by existing technologies, leading to low reuse rates and significant plastic waste accumulation.
Innovation Solution
A computer-implemented method using a database with markers to identify and sort plastic compounds based on their properties, including polymer type, grade, and sustainability score, utilizing chemical tracers, QR-data, and blockchain technology for transparent and secure data management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sorting methods are used for plastic waste, then the sorting process is simple, but the sorting accuracy is low and cannot effectively distinguish between different polymer types, grades, blends, and additives
Solution Approach 1:
The sorting process is segmented into multiple independent stages: initial sorting by polymer type using NIR spectroscopy, followed by secondary sorting by grade and properties using AI image recognition and spectral analysis. This segmentation allows each stage to focus on specific identification tasks, improving overall sorting accuracy without requiring a single overly complex system.
Solution Approach 2:
The system performs preliminary sorting by polymer type before detailed classification by grade and composition. This preliminary action reduces the complexity of subsequent sorting stages by pre-grouping materials, allowing the AI system to focus on finer distinctions within already-separated polymer categories.
2Productivity
If manual sorting methods are used, then the system complexity is low, but the productivity is low and time consumption is high
Solution Approach 1:
Manual mechanical sorting is replaced with an automated system combining NIR spectroscopy for polymer identification, AI-based image recognition for visual defect detection, and automated sorting mechanisms. This substitution dramatically increases productivity while the modular architecture manages system complexity through specialized subsystems.
Solution Approach 2:
The system incorporates self-service features through automated quality assessment and sorting decisions made by the AI algorithm without human intervention. The system automatically adjusts sorting parameters and makes real-time decisions, maintaining high productivity while reducing the need for complex human-machine interfaces.
3Reliability
If sorting is performed without considering sustainability metrics, then the process is simpler, but the economic value and sustainability of recycled materials are reduced
Solution Approach 1:
The system incorporates feedback loops where sorting results are continuously analyzed against sustainability metrics and quality standards. The AI algorithm learns from sorting outcomes and adjusts classification criteria to improve both material quality and sustainability performance. This feedback mechanism ensures high reliability of recycled material while managing data complexity through iterative optimization.
Solution Approach 2:
The sorting system serves multiple functions simultaneously: it sorts by polymer type, identifies grades, detects contaminants, assesses sustainability metrics, and generates quality certificates. This multi-functionality is achieved through a unified AI platform that processes multiple data streams, reducing the need for separate specialized systems and managing overall data complexity.
Data Source
AI summary
A computer-implemented method can be used for controlling sorting of plastic compounds. The method involves providing a computer-based database containing entries on a plurality of markers each identifying a specific plastic compound; receiving scan data from a sample of a plastic compound; identifying a marker in the sample of the plastic compound based on the received scan data and the plurality of markers of the database; and sorting of plastic compounds based on the identified marker of the sample of the plastic compound.