Title: Exploring Robotic Locomotion Modalities on Granular Terrain
Abstract:
Navigating granular surfaces is essential for planetary robotics and for exploring arid regions on Earth for scientific measurements as well as agricultural purposes. Specifically, locomotion on granular slopes poses challenges due to slipping and tilting, often caused by media yielding and downhill flow. To enhance our understanding of granular locomotion, we conducted systematic experiments and utilized simulation tools to investigate various methods of navigating granular terrain. In walking experiments, we examined a planar biped robot on granular slopes, revealing insights into granular physics and effective locomotion strategies. Cleated feet are introduced to enhance traction on steeper slopes. To investigate swimming-like locomotion, we utilized a new class of lab-scale rover robot capable of effectively manipulating flowable slopes, suggesting alternative gaits for planetary rovers. The rover’s locomotion abilities on granular slopes are examined utilizing machine learning techniques and systematic investigations, resulting in optimized gaits. Furthermore, we investigated the subterranean navigation of soft-growing robots using models that mimic plant root growth to unveil robotic anchoring principles. Simulation studies demonstrate the significant role of hair-like protrusions in increasing anchoring forces, offering insights applicable to both robotics and plant science.
Bio:
Deniz Kerimoglu received his Ph.D. degree in Electrical and Electronics Engineering from Bilkent University in 2017. He was a postdoc in Smart Systems Lab at Texas A&M University at Qatar between 2019 and 2021. He joined the Complex Rheology and Biomechanics Lab at Georgia Institute of Technology in 2021 and currently conducts his studies as a research scientist. His research focuses on discovering principles of robotic locomotion on complex terrains by developing novel control approaches and terrain manipulation methods.