Liebherr has introduced a proprietary artificial intelligence system designed specifically for its large-scale excavators, shifting the focus of machine automation from simple grade control to dynamic operational optimization. While current 3D machine control systems excel at telling an operator where to dig to achieve a specific design surface, they do not address *how* to dig efficiently. Liebherr's new AI assistant actively analyzes soil conditions and modifies the machine's hydraulic parameters in real-time to maximize the bucket fill factor and minimize energy waste.
The system relies on a combination of inertial measuring units (IMUs) on the boom and stick, coupled with a forward-looking LiDAR sensor mounted on the cab roof. As the operator begins a digging pass, the LiDAR scans the muck pile ahead, creating a topological map of the material. The AI uses this data to identify the optimal attack angle for the bucket. Simultaneously, the IMUs measure the micro-vibrations and resistance forces acting on the boom during the dig. By analyzing these forces, the system's machine learning algorithm can determine the density and cohesion of the soil-for example, differentiating between loose gravel and compacted clay-and adjust the hydraulic flow curves accordingly.
If the AI detects that the bucket is encountering dense, sticky clay, it will automatically reduce the crowd speed while increasing the breakout force, preventing the bucket from stalling or tracking off-course. Conversely, in loose sand, it speeds up the swing and curl functions to capitalize on the fast-cutting material, ensuring the bucket reaches maximum capacity before the boom reaches the top of the cut. This real-time tuning ensures that every cycle moves the maximum possible volume of material without overloading the engine or causing hydraulic cavitation.
During a controlled test at a major European infrastructure project, a Liebherr R 976 equipped with the AI system achieved a 12% improvement in tonnes moved per hour compared to a manually operated unit working the same material. Furthermore, fuel consumption per tonne dropped by nearly 9%, as the AI eliminated the inefficient over-revving and partial bucket loads typical of manual operation. Liebherr intends to offer this as a software subscription for existing fleets, signaling that the next frontier in heavy equipment productivity isn't just bigger engines, but smarter algorithms.
