Autonomous Excavation of Hydraulic Excavator Based on Real-Time Estimation of Macroscopic Properties of Soil

2022 Fiscal Year Final Research Report

• Project/Area Number: 20H02109
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 20020:Robotics and intelligent system-related
• Research Institution: The University of Tokyo
• Principal Investigator: Nagatani Keiji 東京大学, 大学院工学系研究科(工学部), 特任教授 (80314649)
• Co-Investigator(Kenkyū-buntansha): 谷島 諒丞 東京大学, 大学院工学系研究科(工学部), 特任助教 (10869598) ; 石上 玄也 慶應義塾大学, 理工学部(矢上), 准教授 (90581455) ; 濱崎 峻資 東京大学, 大学院工学系研究科(工学部), 特任助教 (10849003)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Keywords: 自動掘削 / 建設機械

Outline of Final Research Achievements

This study proposed a method for estimating soil properties and generating automatic excavation operations for hydraulic shovels based on the estimated properties. First, a machine learning framework was constructed to predict the time series changes in soil moisture inside the soil from surface images. The framework demonstrated that it could classify four soil types with 90% accuracy. Additionally, it was confirmed that the prediction accuracy could be improved by adding ambient temperature as an input parameter. Next, a proposal was made for automatic excavation operation planning/replanning using the macroscopic properties and shape of the soil. An efficient excavation path was proposed based on the soil density estimated from 3D point cloud measurement of the excavation target and optimization algorithm using a dynamic simulator. The usefulness of the above method was confirmed through simulations and practical experiments with an actual construction excavator.

Free Research Field

フィールドロボティクス

Academic Significance and Societal Importance of the Research Achievements

熟練の油圧ショベルオペレータは，掘削を行う際，土砂の形状のみならず，無意識のうちに土砂の性質を逐次推定し，掘削動作に反映している．このような掘削動作の自動化を機械で実現するため，本研究では，土砂の性質の推定手法と油圧ショベルの自動掘削動作の生成手法を提案し，シミュレーションならびに実建設機械を用いて，その有用性を確認した．このような動作を機械に自動で行わせることは，対象となる土砂が変形するために大変困難であり，ロボットの自律動作獲得ためのマイルストーン動作の一つである．また，同時に，現在の社会問題の解決に直接結びつくことが期待できる結果であるとも言える．