KAUST is developing a robotic system for automated date palm harvesting, combining robotics and AI. The system uses robotic arms with visual sensors to identify and harvest dates, flowers, and tree structures. Field trials are scheduled for the 2025 harvest season, with full operational capability expected within three years. Why it matters: This innovation could transform Saudi Arabia's date farming industry, increasing yields, reducing labor risks, and positioning the country as a leader in agricultural technology.
This paper introduces a convolutional transformer model for classifying tomato maturity, along with a new UAE-sourced dataset, KUTomaData, for training segmentation and classification models. The model combines CNNs and transformers and was tested against two public datasets. Results showed state-of-the-art performance, outperforming existing methods by significant margins in mAP scores across all three datasets.
MBZUAI Professor Fakhri Karray and colleagues from the University of Waterloo are using AI to forecast crop yields, focusing on the impact of extreme temperatures on California strawberry yields. The research uses historical climate and agricultural data to predict yields, addressing issues from 2023 when unusual weather caused a $100 million loss to the strawberry industry. Better predictions could benefit consumers, farmers, and the agricultural industry by improving pricing and supply chain management. Why it matters: This research can improve understanding of agricultural system vulnerabilities amid climate change and extreme weather.
Red Sea Farms, a KAUST startup, is advancing its saltwater greenhouse technology with a new 21,000 square foot pilot facility at the KAUST Research & Technology Park. Their greenhouse technology allows for growing crops on marginal land, using 90% less freshwater than traditional methods. The system uses saltwater in greenhouse-cooling and climate control, resulting in a lower environmental footprint. Why it matters: This technology addresses critical food and water security challenges in arid regions by enabling local food production with minimal freshwater resources and reduced energy consumption.
KAUST's Center of Excellence for Sustainable Food Security (CoE-SFS) has launched 12 translation projects focused on plant growth and water security, establishing partnerships with public and private entities to scale up research. Mark Tester's team developed stress-tolerant rootstocks, grafted onto crops like tomatoes, that thrive in hot, dry conditions with increased yields. Through his start-up Iyris, Tester is conducting commercial field trials in over 12 countries. Why it matters: These efforts to adapt agriculture to environmental change are crucial for ensuring food security in Saudi Arabia, the region, and globally, especially in the face of climate change and limited water resources.
Researchers introduce TomFormer, a transformer-based model for accurate and early detection of tomato leaf diseases, with the goal of deployment on the Hello Stretch robot for real-time diagnosis. TomFormer combines a visual transformer and CNN, achieving state-of-the-art results on KUTomaDATA, PlantDoc, and PlantVillage datasets. KUTomaDATA was collected from a greenhouse in Abu Dhabi, UAE.
This paper proposes a machine learning method for early detection and classification of date fruit diseases, which are economically important to countries like Saudi Arabia. The method uses a hybrid feature extraction approach combining L*a*b color features, statistical features, and Discrete Wavelet Transform (DWT) texture features. Experiments using a dataset of 871 images achieved the highest average accuracy using Random Forest (RF), Multilayer Perceptron (MLP), Naïve Bayes (NB), and Fuzzy Decision Trees (FDT) classifiers.
Researchers in Saudi Arabia have developed a deep learning framework for automated counting and geolocation of palm trees using aerial images. The system uses a Faster R-CNN model trained on a dataset of 10,000 palm tree instances collected in the Kharj region using DJI drones. Geolocation accuracy of 2.8m was achieved using geotagged metadata and photogrammetry techniques.