GhostAR: A Time-space Editor for Embodied Authoring of Human-Robot Collaborative Task with Augmented Reality

Engineers at Purdue University have developed an algorithm known as GhostAR that utilizes motion capture from human movement to create precise task sequencing for robots. The program enhances robotic manufacturing as well as development of human-robot collaboration. Programming a robot through augmented reality (AR) to perform tasks often leads to slow human-robot collaboration, issues with input inaccuracy from imaging to mimic tasks in situ, and creates reliance on editing offline. Human-computer interactions are made more natural through GhostAR's self-contained interface, which allows a machine to learn by robust motional mapping. The system allows for authoring work flows by using motion capture images as input to guide tasks and to ensure constant human-robot awareness. An interactive visual simulation is displayed to the user, which makes it easier to program robots as well as gives better feedback to robots without having to operate offline.

Advantages:

-Robust mimicking capabilities

-Enhanced authoring interfaces

-Heightened robotic awareness

-Realistic simulation and visualization

-Real-time iterative feedback loop

-User-friendly

-Efficient workflow

Potential Applications:

-Manufacturing

-Robotics

-Training

TRL: 2

Intellectual Property:

Keywords: GhostAR, human-robot collaboration, augmented reality in manufacturing, robotic task sequencing, motion capture for robots, embodied authoring, real-time iterative feedback loop, human-robot interaction, AR robot programming, efficient workflow in robotics, Algorithm, Augmented Reality, Collaboration, Computer Programming, Computer Programs, Computer Technology, interaction Styles, Interactions, Interactive, Mechanical Engineering, Robot Navigation, Robotic Task Planning, Robotics, Sequential Task Authoring, Spatial Interactions, Spatial Mapping, Spatial Visualization