Artificial Intelligence Applications in Fruit Production, Harvesting and Postharvest Quality: A Review
Ashutosh Kumar
Dr. Rajendra Prasad Central Agricultural University Pusa Samastipur, ACTO (SMS Horticulture) KVK, Narkatiyaganj, RPCAU, Pusa, India.
Ramawatar Bajia
Faculty of Agriculture, Jagannath University, Jaipur- 303901 (Rajasthan), India.
Preeti Toppo
MGUVV, Durg, College of Horticulture and Research Station, Ramanujganj, 497220, India.
Tejraj Singh Hada
Balwant Vidyapeeth Rural Institute, Bichpuri, Agra, Uttar Pradesh, India.
Shah Ahad Akhtar Khan
Department of Agriculture, Integral Institute of Agricultural Science and Technology, Integral University Lucknow-226026, Uttar Pradesh, India.
Manjul Jain
School of Agriculture, Eklavya University, Damoh, Madhya Pradesh, India.
E. Karuna Sree
Dr. YSR Horticultural University, Andhra Pradesh, India.
Devi Darshan Tiwari *
Department of Agriculture, Integral Institute of Agricultural Science and Technology, Integral University Lucknow-226026, Uttar Pradesh, India.
*Author to whom correspondence should be addressed.
Abstract
Artificial intelligence (AI) is increasingly being explored as a tool for improving decision-making across fruit production and postharvest management. However, evidence regarding its applications, reported performance, and practical limitations remains distributed across different fruit crops, sensing technologies, and computational approaches. This review systematically examines the applications of machine learning, deep learning, computer vision, robotics, the Internet of Things (IoT), and remote sensing in fruit science. Relevant publications were identified through searches of [databases used] covering the period from [start year] to [end year], using predefined search terms and eligibility criteria. Of the [number] records identified, [number] studies were included in the final synthesis. The selected literature was categorised into precision orchard management, pest and disease detection, fruit detection and yield estimation, robotic harvesting, non-destructive quality assessment, and postharvest handling. The synthesis compares the reported performance of AI models, their application across fruit crops, the sensing technologies employed, and the principal limitations affecting field implementation. Particular attention is given to quantitative outcomes reported for detection accuracy, prediction performance, harvesting efficiency, resource use, and fruit quality assessment. The review also evaluates challenges associated with dataset availability, model generalization, computational requirements, sensor integration, and deployment under variable orchard conditions. Emerging approaches, including edge AI, multimodal data fusion, explainable AI, and autonomous orchard systems, are discussed in relation to the limitations identified in existing studies. By connecting reported experimental outcomes with implementation constraints, this review identifies evidence-based research priorities for developing reliable and scalable AI applications in fruit production and postharvest management.
Keywords: Artificial intelligence, precision horticulture, machine learning, deep learning, computer vision, orchard monitoring, remote sensing, robotic harvesting, fruit quality assessment, postharvest management