This project addresses the challenge of monitoring and predicting the behavior of granular materials, such as grains and pellets, that are processed in large volumes across the chemical, agricultural, and mining industries. The mechanical behavior of these materials in industrial operations is highly unpredictable due to external factors such as temperature and humidity, and internal factors such as particle degradation and fracture. Obtaining data is challenging in large industrial systems because they often span tens of meters and have strong spatial variabilities. The lack of predictive capability leads to significant economic losses and safety risks, including silo collapse and personnel entrapment during inspection procedures. The proposed research aims to develop robots with the ability to navigate through large granular systems, much like how animals burrow through soils. The robots will collect critical data and inform large-scale data-driven models that predict system responses to changes in operating conditions. The research will advance fundamental understanding of granular materials while introducing an innovative approach to enhance the energy efficiency, reduce waste, ensure safety, and benefit the welfare of the broader society. Collaboration between academia and industry will facilitate the transfer of fundamental science to industrial applications and offer hands-on experience for students at various academic levels. The technical goal of the project is to