Max Robotics — Global Service Robot Directory + US Certification

About

Dan Gutfreund is a leading researcher at the intersection of embodied AI, computer vision, and robotics, with a focus on building physically realistic and socially intelligent agents. His most impactful work centers on the **ThreeDWorld Transport Challenge**, a benchmark that pushes the boundaries of visually guided task-and-motion planning. This benchmark requires agents to navigate simulated home environments, manipulate objects with articulated arms, and solve complex transport tasks under realistic physics—a contribution that has garnered over 39 citations across its iterations. Gutfreund has also pioneered **asynchronous audio-visual integration**, demonstrating how agents can locate fallen objects by combining sound and sight, a key step toward robust perception in dynamic settings. His work on **probabilistic inverse graphics** introduces uncertainty modeling into 6D pose estimation, enhancing robustness in 3D scene understanding. Beyond physical interaction, Gutfreund explores **rich social interactions** by formalizing microsociological theories into nested Markov decision processes, enabling robots to reason about others' goals. Through these contributions, he is shaping the next generation of embodied agents that can perceive, reason, and act in both physical and social worlds.