Socially CompliAnt Navigation Dataset (SCAND): A Large-Scale Dataset of Demonstrations for Social Navigation

Social navigation is the capability of an autonomous agent, such as a robot, to navigate in “socially compliant” manner presence other intelligent agents humans. With emergence autonomously navigating mobile robots human-populated environments (e.g., domestic service homes and restaurants food delivery on public sidewalks), incorporating socially compliant behaviors these becomes critical ensuring safe comfortable human-robot coexistence. To address this challenge, imitation learning promising framework, since it easier for humans demonstrate task social rather than formulate reward functions that accurately capture complex multi-objective setting navigation. The use inverse reinforcement robots, however, currently hindered by lack large-scale datasets robot demonstrations wild. fill gap, we introduce Socially CompliAnt Navigation Dataset ( SCAND )–a large-scale, first-person-view dataset demonstrations. Our contains 8.7 hours, 138 trajectories, 25 miles compliant, human tele-operated driving comprises multi-modal data streams including 3D lidar, joystick commands, odometry, visual inertial information, collected two morphologically different robots–a Boston Dynamics Spot Clearpath Jackal–by four demonstrators both indoor outdoor environments. We additionally perform preliminary analysis validation through real-world experiments show policies learned generate behaviors.