Optimizing wheel loader performance: an end-to-end approach

2025-01-11

AI Summary

To address efficiency limitations in continuous loading tasks for wheeled loaders operating in dynamic, unstructured mining and construction environments—where evolving stockpile configurations hinder performance—this paper proposes an end-to-end optimization framework. The method integrates a deep neural network-based world model with a lookahead multi-step tree search to jointly model the spatiotemporal evolution of stockpile states and the 15-step loading sequence. A novel V-cycle transportation cost function is introduced to enable global objective optimization. Compared to conventional one-step greedy policies and calibrated fixed controllers, the proposed approach improves overall transport efficiency by 6% and 14%, respectively, over 15 consecutive loading steps. It overcomes inherent limitations of local optimization and static control strategies, establishing a new paradigm for autonomous decision-making in dynamic, non-structured heavy machinery operations.

Technology Category

• Search and Optimization: Learning to Search
• Intelligent Robots: Learning & Optimization for ROB
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Wheel loaders in mines and construction sites repeatedly load soil from a pile to load receivers. This task presents a challenging optimization problem since each loading's performance depends on the pile state, which depends on previous loadings. We investigate an end-to-end optimization approach considering future loading outcomes and V-cycle transportation costs. To predict the evolution of the pile state and the loading performance, we use world models that leverage deep neural networks trained on numerous simulated loading cycles. A look-ahead tree search optimizes the sequence of loading actions by evaluating the performance of thousands of action candidates, which expand into subsequent action candidates under the predicted pile states recursively. Test results demonstrate that, over a horizon of 15 sequential loadings, the look-ahead tree search is 6% more efficient than a greedy strategy, which always selects the action that maximizes the current single loading performance, and 14% more efficient than using a fixed loading controller optimized for the nominal case.

Problem

• Wheel Loader Optimization
• Dynamic Soil Condition
• Operational Efficiency

Innovation

• World Model Integration
• Tree Search Algorithm
• Dynamic Optimization in Excavation