Dr. Shukhrat Shokirov
Theme C: Data-Driven Solutions – Leveraging AI and Big Data
09.10.2025 09:45 – 10:30
Conference Hall #1
Moderator: Lars Ribbe
Dr. Shukhrat Shokirov is the Vice-Rector for Research and Innovation at TIIAME National Research University in Tashkent, Uzbekistan. He holds a PhD from the Australian National University and an MSc in Geoinformatics from the University of West Hungary. His research focuses on remote sensing, LiDAR, and geospatial analysis for vegetation monitoring, forest structure assessment, and precision agriculture.
Dr. Shokirov has held postdoctoral positions at the University of Maryland and ANU, contributing to international projects on satellite-based ecosystem monitoring and yield prediction. He has published in leading journals, such as Remote Sensing of Environment and the International Journal of Remote Sensing. He has been invited as a keynote speaker at conferences in China, Uzbekistan, and the USA. His current work integrates machine learning and cloud platforms, such as Google Earth Engine, to support data-driven agricultural innovation in Uzbekistan and beyond.
Image © Private | Dr. Shukhrat Shokirov
Abstract
Space-Based Monitoring for Sustainable and Data-Driven Crop Production
The fusion of big data and cloud-based geospatial platforms is revolutionizing precision agriculture by enabling scalable, efficient, and timely decision-making. In this study, we leverage the computational power of Google Earth Engine (GEE) to extract and process multi-temporal Sentinel-2 imagery for the monitoring of winter wheat growth and yield prediction. GEE’s cloud-native architecture allows seamless access to petabyte-scale Earth observation archives and facilitates the calculation of vegetation indices such as NDVI, NDRE, GNDVI, and EVI across large agricultural areas in just minutes with minimal code. This substantially reduces data acquisition and preprocessing time compared to conventional desktop-based workflows.
By combining satellite-derived spectral indicators with ground-based yield measurements, we trained several machine learning models to assess crop performance and forecast yield with high accuracy. Among them, the Gradient Boosting model demonstrated superior predictive capability. The approach supports the early detection of crop stress, disease, and nutrient deficiencies, empowering farmers and land managers to take timely and targeted interventions. Additionally, it reduces reliance on costly and labor-intensive field surveys by offering continuous, remote crop monitoring capabilities.
This study illustrates the transformative potential of integrating remote sensing, machine learning, and cloud computing in the agricultural sector. It not only enhances the precision of crop management decisions but also promotes resource efficiency and long-term sustainability. As the methodology is scalable and adaptable, it holds promise for broader applications in other crops and regions, especially in data-scarce or resource-limited environments. Ultimately, this work contributes to the advancement of data-driven agriculture and the development of resilient food production systems in the face of global environmental challenges.