Hybrid Robotic Weeding Systems: An Evidence-Qualified Systematic Review of Taxonomy, Performance, and Context-Aware Adaptation
Open Access
Abstract
Robotic weeding has emerged as an important technology in precision agriculture. However, existing reviews rarely distinguish integrated hybrid architectures from mono-method robotic systems or sensing-only platforms. This lack of distinction limits reproducible classification and evidence synthesis. This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 and the PRISMA extension for Searches. The literature search covered Scopus, Web of Science, ScienceDirect, SpringerLink, and Wiley Online Library. Studies published between January 2011 and August 2026 were considered. A total of 248 records were identified. Of these, 109 studies met the eligibility criteria and were classified according to the directness and maturity of evidence for hybrid-system integration. An evidence-qualified two-level taxonomy was developed based on physical coupling and spatial integration. Hybrid-system performance was then synthesised. The synthesis also considered the biological, environmental, and operational determinants of context-aware adaptation. Mechanical–chemical systems currently provide the strongest direct field evidence. Their weed-control effectiveness reaches up to 99.5%, with herbicide reductions of up to 65.6%. In contrast, mechanical–laser systems are supported mainly by prototype evidence, while the remaining coupling categories are still provisional or insufficiently validated. Overall, the available evidence is too heterogeneous to support universal performance rankings. This evidence-qualified framework can support R&D prioritisation, system-architecture selection, and engineering decision-making by distinguishing validated hybrid configurations from provisional concepts and by clarifying the evidence required for modality coupling, spatial integration, and field validation.Keywords
Funding Information
- Science Committee of the Ministry of Science and Higher Education of the Republic of Kazakhstan (AP23490705)
This publication has 125 references indexed in Scilit:
- A Field‐Adaptive Mechanical Weeding System Coupling Oscillating Pneumatic Mechanism With Deep Learning for Intra‐Row Weed Control in LettuceJournal of Field Robotics, 2026
- Deep learning‐based laser weed control compared to conventional herbicide application across three vegetable production systemsPest Management Science, 2025
- Sensor-Guided Mechanical Weed Control in Transplanted Lettuce and CabbageJournal of Crop Health, 2025
- Field efficacy of a bioherbicide mimic (DiS‐NH2) in a nanoparticle formulation for weed control in durum wheatPest Management Science, 2025
- Applications of High-Power Semiconductor Lasers in the Field of Ecological Environment Detection and Protection2024
- Evaluation of two deep learning‐based approaches for detecting weeds growing in cabbagePest Management Science, 2024
- Application technology for bioherbicides: challenges and opportunities with dry inoculum and liquid spray formulationsPest Management Science, 2023
- Power on! Low‐energy electrophysical treatment is an effective new weed control approachPest Management Science, 2021
- Technology readiness levels: Shortcomings and improvement opportunitiesSystems Engineering, 2020
- Robot for weed species plant‐specific managementJournal of Field Robotics, 2017