Developed in collaboration with analytics expert RTT System GmbH, this company has unveiled the BRT HARTNER AS Automatic Sorter, its inaugural sensor-based and AI-enhanced separation platform designed for high-precision sorting of materials that conventional methods cannot process.

Source: Chemengonline

The AI is capable of processing all recorded data without loss, enabling it to reliably identify materials and categorize objects into predefined classes. This capability allows for precise assessments about materials, ensuring that various types of plastics can be accurately distinguished for recycling.

The BRT HARTNER AS integrates an acceleration belt, optical sensors, and compressed-air sorting. Unlike many sorting systems that depend on near-infrared (NIR) technologies, the AS system features the latest hyperspectral imaging (HSI) technology, as highlighted by the manufacturer. This HSI camera system not only produces high-resolution images but also delivers vital information regarding material properties.

Moreover, the BRT HARTNER AS system is distinguished from other market solutions by its sensor fusion-based concept. With sensor fusion, the raw data from all installed sensors (NIR/color/3D/metal) is evaluated directly, eliminating additional interfaces. Consequently, the data from all sensors is transmitted for analysis without any loss. The detection algorithm has been trained on extensive datasets, demonstrating remarkable precision. Users can view the material classification of the input material in real-time on the system’s user-friendly display. For instance, in the case of lightweight packaging, the system not only indicates the material type with color coding but also captures and records the specific packaging type.

The system's versatility allows for the seamless integration of new sensors at any time at the raw data level, enabling the retrofitting of 3D sensors, metal sensors, or future marker technologies. If needed, the standard spectral AI can also be enhanced with object AI (a color camera must be installed for this feature), which can precisely identify and segregate specifically defined objects.