BROWSE
Detail
Conference Paper
RoboManuGen: Robot Data Generation using Generative AI for Manufacturing
Cited 0 time in 
Share 
Share
- Authors
- Hyejin S. Kim
- Issue Date
- 2024-10
- Citation
- International Conference on Information and Communication Technology Convergence (ICTC) 2024, pp.1-4
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- This paper introduces RoboManuGen, a robot generation simulator designed to create training data for various manufacturing robots through large-scale generative simulation. Built on NVIDIA's Issac Sim, RoboManuGen encompasses diverse robots, various grippers, a range of manipulation tasks, manufacturing environments, and key objects handled within these settings. It also simulates anomalous behaviors of robots and anomalies in objects occurring during manufacturing processes for anomaly detection in a robot manufacturing process. In the future, leveraging RoboManuGen, we intend to develop generative models that automatically generate diverse tasks, scenes, and training supervisions using this data.
- KSP Keywords
- Data generation, Generative models, Manufacturing processes, anomaly detection, large-scale, training data