Development of ultra-precision weeding technology using a system for unmanned-ground-vehicles driven by solar power

2023 Fiscal Year Final Research Report

• Project/Area Number: 21H02321
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 41040:Agricultural environmental engineering and agricultural information engineering-related
• Research Institution: Akita Prefectural University
• Principal Investigator: Satoshi Yamamoto 秋田県立大学, 生物資源科学部, 准教授 (20391526)
• Co-Investigator(Kenkyū-buntansha): 保田 謙太郎 秋田県立大学, アグリイノベーション教育研究センター, 准教授 (00549032) ; 間所 洋和 岩手県立大学, ソフトウェア情報学部, 教授 (10373218) ; 西村 洋 秋田県立大学, アグリイノベーション教育研究センター, 特任教授 (70391513)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: 農業ロボット / 自律走行 / 自動除草 / 三次元再構築

Outline of Final Research Achievements

We developed an autonomous robot equipped with solar panels to perform automated weeding of soybean and buckwheat, and it was confirmed that the robot could repeatedly move through a 1-hectare field. However, it was not possible to establish a method for timely weeding because the operating speed was not sufficient compared with the growth speed of weeds. Therefore, a battery-powered radio-controlled mower was modified to perform automatic weeding in a soybean field. As a result, while the weeding was reliable, the soybean leaves were entangled and damaged, suggesting the need for a weeding mechanism that is both efficient and accurate. For the weed detection technology, a semantic segmentation model was constructed to improve segmentation accuracy from a small amount of image data and labels.

Free Research Field

農業機械

Academic Significance and Societal Importance of the Research Achievements

ロボット農機の完全無人化について盛んに研究されているが、周囲の作業者や障害物を確実に検知する安全性確保が課題である。本研究では、軽量・低コストなエネルギー自給型の自律走行ロボットにより、安全性が確保された完全無人の除草体系の確立を目指した。除草の作業適期への対応が困難であったが、小型ロボットをほ場で運用するノウハウや、深層学習による雑草検出、ほ場の三次元再構築について成果を得ることができた。