Kobelco Construction Machinery has launched the "Geo-Link Smart" intelligent bucket-fill optimization system for its SK350LC-11 series 35-ton excavators, targeting quarry stripping and bulk overburden excavation where bucket fill factor varies dramatically with material type. In a typical face-loading cycle, the operator digs through alternating layers of soft clay, weathered shale, and competent limestone. Traditional excavators rely on the operator's visual assessment to adjust dig depth and crowd speed. An operator digging too shallow in soft material fills the bucket to only 60 percent capacity, while digging too deep in competent rock stalls the machine and wastes cycle time.
The Geo-Link Smart system integrates "High-Resolution Pressure Transducers" on the boom, arm, and bucket cylinders, combined with a "Swing Torque Sensor" on the upper structure. The engineering innovation is the "Fill-Factor Signature Recognition." During each dig cycle, the system records the pressure-versus-time curve from all three cylinders. This signature encodes the material's resistance profile-soft clay produces a smooth, low-pressure curve with high arm-crowd speed, while competent limestone produces a jagged, high-pressure curve with slow crowd. The onboard AI classifier has been trained on thousands of dig cycles across multiple material types and identifies the material being excavated within the first 500 milliseconds of the dig.
Based on the classified material, the system preemptively commands the "Optimal Dig Trajectory"-adjusting the arm's crowd rate and the bucket's curl timing to maximize payload. In soft material, the system deepens the dig and increases crowd speed to fill the bucket faster. In hard material, the system shallows the dig to prevent stall and increases bucket curl force to shear the material into the bucket. The system also monitors the swing torque when the loaded bucket rotates to the haul truck; if the swing torque indicates an under-filled bucket, the system adjusts the next dig cycle's parameters accordingly. Across a full shift, this automated fill optimization increases bucket payload per cycle by an average of 12 percent and reduces fuel consumed per ton of material moved by 9 percent.
