Welcome to ROS Embodied AI Community Group! Please refer to our living document for meeting link, notes, planning and agenda.
The group unites the open source community developing Generative AI frameworks, tools and applications for ROS ecosystem. It serves to align developments and as a platform to share experiences with novel technology, learning from other projects and applications, and collaborating towards common goals, new features, and integrations. We are also looking forward towards standardization efforts as the technology and projects mature.
Adam Dabrowski (Robotec.ai), adam.dabrowski@robotec.ai, GitHub handle @adamdbrw.
Large Language Models, Vision Language Models and AI Agent architectures are recent breakthrough technologies that most of you are familiar with. There are many gaps in Generative AI for robotics such as in spatial and temporal reasoning, dealing with imperfect perception and actuation, performance, size, liability, and so on. However, there are also important strengths and applications that are already quite reasonable, such as in human-robot interaction, high-level planning, inspections, lowering the skill gap of robot use, and diagnostics.
So far, the following project leads expressed interest in participating:
- https://ros-wg-embodied-ai.github.io/resources - Github Pages for the ROS Embodied AI Community Group links and resources, including talks, demos, group logos, etc.
- awesome-embodied-ai-ros - the curated list of repositories and papers on Embodied AI in ROS ecosystem.