Results 1 - 1 of 1
Results 1 - 1 of 1. Search took: 0.018 seconds
[en] Some of the global challenges currently facing the food and agriculture sector are so multifaceted that they cannot readily be solved by human expert knowledge alone. As AI and ML techniques mature, the opportunities to implement these new methods in numerous scenarios within the domain of nuclear techniques in food and agriculture will arise. Applications may include food fraud detection, predicting food safety incidents, remote sensing data for agricultural soil management, optimising remediation of radioactively contaminated land, and development of new food and beverage products. Extensive information at all scales is needed in decision-making processes for agriculture and food production. Enhanced data availability through the implementation of open data policies and innovative data acquisition methods, have enabled the use of AI in the food and agriculture sector. Variation in sampling, sample preparation and analysis are often bottlenecks for data sharing, but AI could assist in dataset standardisation. Furthermore, AI can help by improving analytical prediction based on traditional chemometrics (e.g., infrared spectroscopy, nuclear magnetic resonance), or by supporting calibration of analytical equipment. Regarding the latter, AI can further assist in calibrating measurements carried out across the same type of equipment but from different providers, essentially leading to the creation of large datasets needed for AI applications. In addition, AI can play an essential role in bringing data together from different sources, resolutions or scales, and lastly, the Internet of Things combined with AI and decision support is a fast-growing domain. However, legal constraints and ethical concerns around security, trust and issues of transparency and explainability are limiting the applications of AI methodologies in food and agriculture, as in other fields. It is important to remember that even with the full potential of AI unleashed within the food and agriculture domain, data-driven decision-making is only as good as the data used. Properly coupling available data requires state-of-the-art and validated solutions that focus on harmonising data access, analytics and predictive modelling.