ABU DHABI, 28th February, 2022 (WAM) — The Abu Dhabi Agriculture and Food Safety Authority (ADAFSA) in collaboration with the Abu Dhabi Digital Authority and Emirates ICT Innovation Centre (EBTIC) has successfully developed a model for UAE Food and Animal Feed Import Prediction using artificial intelligence and machine learning.
This model will build a vision of the direction and value of food imports to the UAE, and analyse the most imported strategic foodstuffs, their sources and quantities to identify potential diversification opportunities for imports and reduce waste.
The model was built by developing an interactive pictorial dashboard by analysing a series of historical data on the quantities and values of imports and re-exports of animal, agricultural and feed food products. The panel shows interactive information on countries from which food and feed are imported, the quantities and values to create a machine learning model to predict the amount of plant and animal food and animal feed imported, and determine the size and cost of imports and the countries from which imports are imported. The model also analyses food waste, thereby identifying potential opportunities for re-export.
Aysha Al Naili Al Shamisi, Statistics and Analysis Division Director at the Abu Dhabi Agriculture and Food Safety Authority, said that developing a model for predicting food imports and understanding the food and feed import trends helps to develop a clear picture of future levels of food imports. It enables the development of policies, regulations and plans to ensure the continuity of the State’s food supply and assists in developing accurate response plans for disasters and crises, or in the event of any food supply disturbances, by knowing the quantities currently imported.
Shamsi explained that the model’s development was based on data on the country’s food imports from 2015 to 2020. The data on animal food imports was relied upon for products such as eggs, frozen chicken meat, fresh chicken meat, and agricultural products such as cucumber tomatoes, watermelons, feed, and hay, to understand the level of dependence in meeting the needs of consumers for each product and its country of origin. It helped to determine the nature of potential risks in terms of supply, prices and opportunities to find alternatives in times of crises and emergencies or in the event of a lack of supply or high prices. She noted that the model provides detailed information on waste rates for each product within the supply chain, which helps strengthen accounting mechanisms for food waste and create re-export opportunities.
Al Shamsi pointed out that the model allows adding updates for requirements, noting that updates and planned additions to the models developed include the addition of new food products to analyse their data with the same mechanism of analysis of product data currently available. This is in addition to setting criteria for developing the performance of the import forecast model, such as seasonal changes of product data, and a model to predict local production, to determine the amount and value of expected production of local food products, depending on the seasonal changes of products data, with artificial intelligence and machine learning techniques.