Max Robotics — Global Service Robot Directory + US Certification

About

Chad DeChant is a robotics and artificial intelligence researcher whose work spans two compelling frontiers: robotic perception and manipulation, and natural language interfaces for human-robot interaction. He is perhaps best known for his highly influential 2017 paper "Shape Completion Enabled Robotic Grasping," which has accumulated nearly 300 citations and introduced a pioneering architecture that allows robots to plan grasps by inferring complete 3D object shapes from partial observations using convolutional neural networks. To support this work, DeChant and colleagues contributed an open-source dataset of over 440,000 3D exemplars, a resource that has since benefited the broader robotics community. More recently, DeChant has turned his attention to the challenge of robot transparency and explainability, developing systems that enable robots to summarize and answer questions about their own past actions using natural language. His work on episodic memory verbalization, leveraging large language models and hierarchical representations of long-horizon robot experience, addresses a critical gap in human-robot trust and communication. Taken together, his research reflects a consistent drive to make robots both more capable and more comprehensible — equipping them not only to act intelligently in the world, but to explain themselves to the humans they work alongside.