Staer Sponsors Nav2: Extending ROS with Visual Semantic Reconstructions

Staer joins an distinguished list of Nav2 Sponsors committing to the long-term development and maintenance of the industry standard framework. Staer is building the physical AI platform for robust, efficient, and safe fleet coordination for autonomous mobile robots (AMRs). The partnership with Open Navigation seeks to unlock the potential for high-fidelity navigation across diverse environments including homes, warehouses, agricultural fields, and more.

Staer Nav2 integration preview Preview of 3D semantic map integration in Nav2

Staer & Nav2

Through Staer's platform, users will have access to industry-leading visual perception and planning capabilities from 3D semantic mapping to vision-language navigation (VLN) models. Nav2 is the foundational library essential for achieving autonomy for mobile robots in diverse domains. With Staer, Nav2 users will have new and unprecedented access to dense 3D maps for planning and control, containing not only spatial occupancy but also the semantic layer within a scene. This will be integrated into Nav2 standard workflows for navigating scenes with semantic knowledge and for autonomy applications to leverage Spatial AI in their mission planning.

Staer Nav2 integration layer Staer offers shared Spatial Intelligence and Coordination layers

Staer's commitment demonstrates a belief that integrating visual mapping and navigation as core capabilities within Nav2 will empower the community to innovate and accelerate development across a wide range of AMR stacks. This partnership will benefit the Nav2 and ROS communities by putting a focus in the community on how to work with dense and semantically rich 3D environmental models from navigation to manipulation and general scene understanding.

Next Steps

Moving forward, we plan to design and implement Nav2 extensions for visual navigation for robot fleets, utilizing 3D semantic maps (where a fleet of robots collaborate to create a map from camera data and subsequently navigate it). We will follow up with long-term initiatives focused on continuous updates for these 3D semantic maps.

We are excited to collaborate with the community in the open. By working together on the core components in Nav2, we will advance mobile robotics together, without siloed reinvention or reimplementation!