GOALI: Self-propelling robots for the monitoring and data-driven modeling of bulk granular processes

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

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 develop the mechanics for deploying self-propelling robots with sensing capabilities that continuously survey a large granular medium, infer material data representative of the entire system, and predict the behavior of unexplored systems. The team will first utilize 3D experiments, X-ray imaging, and simulations to investigate interaction dynamics between a deformable robot and various granular media. Physics-based analyses correlate the forces and deformations of the robot to the mechanical properties of the surrounding granular medium. This will allow the robot to survey key granular properties as it moves, such as material strains, pressure, and yield stress. Leveraging the data collected by the robots, a data-driven continuum mechanics model will be developed to predict the behavior of the granular system under varying operating conditions. Experimental test beds with embedded robotic sensors will validate this modeling framework. The proposed data-driven approach, combining granular intrusion physics and continuum mechanics, will be an important step for monitoring and simulating industrial-scale systems of complex materials where traditional phenomenological models face notable limitations. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.