Max Robotics — Global Service Robot Directory + US Certification

About

Shun Ogasa is a pioneering roboticist whose work sits at the intersection of tactile sensing, dexterous manipulation, and machine learning. His research focuses on endowing robotic hands with a sense of touch sophisticated enough to recognize objects and perform stable, in-grasp manipulations—a critical challenge for real-world robotics. Ogasa’s most impactful contribution is the integration of his custom **uSkin 3-axis tactile sensors** into a multi-fingered Allegro Hand, enabling dense, triaxial force vector measurements. His landmark 2018 study (45 citations) demonstrated how this rich tactile data dramatically improves object recognition during active sensing. He further advanced the field by applying deep learning architectures—specifically CNNs and CNN-LSTMs—to achieve stable in-hand manipulation with low-cost, imprecise robotic hands (20 and 14 citations, respectively). Notably, his 2020 work on variable in-hand manipulation showed that a single trained model could generalize across diverse tasks like rolling, rotating, and sliding objects, reducing the need for task-specific training. Ogasa has also systematically investigated how fingertip design parameters—such as hardness, skin thickness, and shape—influence prehension stability, providing foundational design principles for the field. His work is essential reading for anyone interested in closing the tactile feedback loop for truly dexterous robots.