Max Robotics — Global Service Robot Directory + US Certification

About

Shane Griffith is a robotics and artificial intelligence researcher whose work centers on interactive object recognition, multimodal sensory learning, and behavior-grounded categorization in robotic systems. His research explores how robots can develop meaningful object representations through direct physical interaction with their environments — an approach deeply inspired by principles of embodied cognition and infant developmental psychology. Griffith's most influential contribution, "Interactive Object Recognition Using Proprioceptive and Auditory Feedback" (2011, 63 citations), demonstrates how robots can identify household objects by performing exploratory behaviors such as lifting, shaking, and crushing, then analyzing the resulting sensory feedback. This work is complemented by his behavior-grounded framework for separating containers from non-containers (47 citations), which elegantly unifies information across multiple sensory modalities — including acoustic and visual signals — to form coherent object categories. Throughout his body of work, Griffith consistently champions the idea that a robot's understanding of objects must be rooted in its own sensorimotor experience rather than imposed through purely symbolic or visual means. His graph-based movement dependency representations further extend this vision by capturing rich interaction possibilities between robotic hands and objects. With over 160 cumulative citations, Griffith's research has meaningfully advanced the field of interactive robot learning and grounded perception.