Artificial intelligence for prediction of shelf-life of various food products: Recent advances and ongoing challenges

Accurate estimation of shelf-life is essential to maintain food safety, reduce wastage, and improve supply chain efficiency. Traditional methods such as microbial and chemical analysis, and sensory evaluation provide reproducible results but require time and labor and may not be suitable for real-time or high-throughput applications. The integration of artificial intelligence (AI) with advanced analysis techniques offers a suitable alternative for rapid, data-driven estimation of shelf-life in dynamic storage environments. The current review assesses the application of AI-based techniques such as machine learning (ML), deep learning (DL), and hybrid approaches in food product shelf life prediction. This study highlights how AI can be utilized to examine data from non-destructive testing methods like hyperspectral imaging, spectroscopy, machine vision, and electronic sensors to enhance predictive performance. The review also describes how AI-based techniques contribute to managing food quality, reduce economic losses, and enhance sustainability by ensuring optimized food distribution and reducing waste. AI techniques overcome conventional techniques by considering intricate, multi-sourced information capturing microbiological, biochemical, and environmental factors influencing food spoilage. Meat, dairy, fruits and vegetables, and beverage case studies illustrate AI techniques' superiority in real-time monitoring and quality assessment. It also identifies limitations such as data availability, model generalizability, and computational cost, constraining extensive applications. Cloud and Internet of Things (IoT) platform integration into future applications has to be considered to enable real-time decision-making and adaptive modeling. AI can be a paradigm-changing tool in food industries with intelligent, scalable, and low-cost interventions in food safety, waste reduction, and sustainability. • AI incorporates non-invasive methods for dynamic, real-time shelf-life forecasting. • Machine vision and spectroscopy improve shelf-life monitoring of fresh produce and meat. • Hybrid AI models enhance shelf-life prediction accuracy under dynamic storage conditions. • AI reduces food waste, energy use, and carbon emissions across supply chains. • Deep learning predicts spoilage by analyzing microbial, chemical, and environmental data.