<![CDATA[Background: Accounting for 85% of worldwide palm oil production, Indonesia holds an unrivaled position as the industry’s dominant global supplier. Despite this prominence, Fresh Fruit Bunch (FFB) grading - the quality critical process of assessing fruit ripeness prior to processing remains largely dependent on manual operator judgment, a practice that inherently generates evaluation bias, output variability, and measurable financial inefficiencies at the mill level. Objective: Employing the PRISMA 2020 protocol as its methodological backbone, this paper conducts a Systematic Literature Review (SLR) to consolidate and critically evaluate available empirical evidence concerning the deployment of artificial intelligence (AI), automation, and digitalization within FFB grading operations. Methods: Three peer-reviewed databases: Scopus, ScienceDirect, and Google Scholar were systematically searched to retrieve publications spanning 2018 to 2025. Upon completion of multistage eligibility screening and quality evaluation using the Mixed Methods Appraisal Tool (MMAT) 2018, a corpus of 29 methodologically sound articles was retained for synthesis, supported by strong inter-rater consistency (Cohen’s Kappa k=0.82), almost perfect agreement). Results: The synthesis converged on four principal findings: (1) CNN and YOLO-based architectures deliver FFB classification accuracy reaching 97%, representing a substantial performance advantage over manual evaluation; (2) three dimensions - Technology Readiness, Organizational Readiness, and Human Capability consistently determine implementation outcomes, with organizational readiness associated with projected adoption rate growth of 15–20% over five years; (3) AI-driven grading systems produce verifiable gains in Oil Extraction Rate (OER) alongside reductions in Free Fatty Acid (FFA) levels, with quantifiable business value materializing within an extended post-implementation period consistent with the IT productivity paradox; (4) despite its methodological superiority for confirmatory theory testing, CB-SEM remains strikingly underemployed in this research domain. Conclusion: Among the most consequential findings of this review is the absence of a unified structural model capable of linking mill readiness dimensions, technology adoption pathways, and downstream business performance outcomes, a gap whose resolution is both theoretically necessary and practically urgent. To address this, the review presents a systematic empirical agenda oriented toward future investigation within the Indonesian palm oil industry context.]]>
Copyrights © 2026
