Kleemann GmbH, a Wirtgen Group brand, has launched the "Grain-Sense" intelligent material classification system for its MOBIRET MR 130iPRO mobile triple-deck screens, targeting aggregate production and construction-and-demolition (C&D) recycling where the feed material's size distribution varies unpredictably. Traditional screens operate at a fixed vibration frequency and amplitude, tuned for an assumed average feed material. When the feed contains a higher-than-expected proportion of fines or oversize material, the fixed screen settings produce poor classification efficiency-either allowing fines to contaminate the oversize product or rejecting usable material into the oversize stream, both of which reduce the plant's overall yield.
The Grain-Sense system integrates a "High-Speed Industrial Camera Array" positioned above the screen's feed conveyor. The engineering breakthrough is the "Real-Time Particle Size Distribution (PSD) Analysis." As the feed material passes beneath the camera array, the system captures 500 frames per second. An onboard machine-learning classifier-trained on thousands of hours of aggregate and recycled concrete imagery-analyzes each frame and calculates the particle size distribution in real-time. The AI recognizes individual particle edges and estimates each particle's three-dimensional volume, producing a live histogram of the feed's composition within 200 milliseconds.
Based on the measured PSD, the ECU commands three simultaneous adjustments. First, the "Screen Deck Vibration Drive" adjusts its stroke frequency and amplitude: finer material requires higher frequency and lower amplitude to promote stratification, while coarser material benefits from lower frequency and higher amplitude to bounce large particles across the deck. Second, the "Feed Conveyor Belt Speed" is modulated to maintain a uniform monolayer of material on the screen deck-overfeeding causes mat blinding and classification loss, while underfeeding wastes screen capacity. Third, the "Product Conveyor Splitter Gate" repositions to redirect material between product bins based on the actual PSD, maximizing the yield of the highest-value product fraction. Across a full production shift, this automated optimization increases product yield by 12 percent and reduces screen blinding incidents by 60 percent, significantly improving plant uptime in variable-feed operations.
