Towards the Autonomous Harvesting of Leafy Greens in Deep-Water Culture Systems: A Systematic Review

Abstract

Robot harvesting in Controlled-Environment Agriculture (CEA) farming remains a bottleneck for producing leafy greens in Deep-Water Culture (DWC) systems. Vertical farming, greenhouse, aeroponics and other such platforms have already yielded several working and near-working harvesting solutions. Since raft-based DWC remains comparatively unaddressed, this review surveys harvesting-relevant technologies. Perception, crop harvesting, gentle handling and harvesting integration must work together. This systematic review critically maps these capabilities in CEA, with potential applications to DWC. The PRISMA 2020 framework retrospectively documents the literature identification process. Eight database exports dated 25 July 2026 yielded 6114 records, amongst which 4884 unique records were screened, and 29 reports advanced to full-text retrieval. A total of 4855 records were excluded at the title/abstract level. The present manuscript includes 19 sources used in the evidence map. The synthesis distinguishes 13 core harvesting studies, 5 contextual CEA automation studies and 1 foundational vision source. The literature shows strong results at the component level. However, the reported outcomes are not directly comparable due to the differing metrics that were used to evaluate each element’s performance. This review therefore illustrates the transition from isolated technical functionalities to integrated automation while showing that mature and comparable harvesting systems for leafy greens are yet to be established.