Hi all, your friendly neighborhood navigator here!
I wanted to let you know Nav2 hits it V1.0 today just in time for release into ROS2 Galactic! That means we have a stable API that we’re not planning on changing in the immediate or near-term future, a high degree of stability in the system, and some really great algorithms we now support exclusively in Nav2:
- Keepout zones
- Speed restricted zones
- Hybrid-A* planning
- Regulated Pure Pursuit control
- Navigate through intermediate pose constraints
- Waypoint following with at-waypoint task executors for custom applications
- Modular control, planning, recovery, and navigator servers
- Fully reconfigurable navigation logic through behavior trees and custom behavior tree nodes
- Lifecycle and bond management for deterministic bringup/shutdown and active server heartbeats
- Rviz2 demo plugin for lifecycle and navigation capabilities with feedback
- Support for circular differential, omnidirectional, and non-circular ackermann and legged robots
- and much more!
I’m excited with all of we have accomplished since Nav2 was started back in 2018. I wanted to give a shout-out to some of the major contributors (on-going and former) that has helped make this possible:
- @amerzlyakov Alexey at Samsung Research
- @mkhansen Matt formerly at Intel
- @bpwilcox Brian formerly at Intel
- @ruffsl UCSD and our fearless CI/CD leader for which our CI system would be broken and I’d be a fumbling idiot trying to fix constantly
- Mohammad Haghighipanah at Intel
- @orduno Carlos formerly at Intel
- Sarthak Mittal from Manipal Institute of Technology & Polybee who’s really stepped up in the last year since joining the project and made substantial high-impact contributions
I’m excited where the next 3 years will take us for our V2.0! Some projects that are on our minds or are actively being worked on now:
- State Lattice planner for non-circular differential / omnidirectional robots
- Dynamic obstacle detection with many sensor modalities, tracking, and avoidance pipelines
- Semantic information integration (rooms, specific objects, or virtual points)
- Retuning / improving DWB / adding critics to have a better out-of-the-box experience
- Additional first-time setup tutorials, tuning guides, and programming examples
- Assisted Tele-operation using behavior trees
- Generic Docking software
- Outdoor capable non-planar environmental models
- Visual and 3D SLAM integrations (with generic localizers for each modality)
- GPS and multi-source global integration demos
- And much more!
While no piece of software is ever 100% perfect, I’m happy that we have over 85% test coverage with full doxygen coverage as well in our V1.0. We expect to continue fielding minor issues from users and resolving them quickly, but its definitely time to stand up and be proud of what we’ve made as a community! If we waited for any project to be 100% perfect, we’d never release a V1 .
So I encourage you if you haven’t tried Nav2 yet to take a look today! The binaries for Nav2 and related projects (slam_toolbox, robot_localization) will be available on Day 1 of the new Galactic release in all of its V1.0 glory. I’m always super happy to see all of the great companies and research that makes use of this work. We have users using Nav2 in warehouses, to pick up trash in parks, startups the world over, massive technology giants, Aerospace, legged robots, vertical farming, and much more! This also includes the TRI Olympics 2021 teleoperation robots for disabled and remote people as well as the Elroy Air massive drone for run-way and ground operations (P.S. if your robot uses Nav2, consider adding it to our Robots Using Page):
If you’d like to get involved with Nav2 please feel free to comment on this, shoot me a PM, or join our Nav2 community slack group. We’d love to have you involved and we can find projects both big and small for any skill level or time commitment. I hope we can continue coming together as a community to build the worlds best mobile robot navigation system in Nav2. Students, professionals, researchers, and hobbyists are all welcome! I definitely could not keep the lights on without all of your help.
If you’re looking for a paper to cite:
- S. Macenski, F. Martín, R. White, J. Clavero. The Marathon 2: A Navigation System . IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020.
Until next time, happy navigating,
Steve Macenski – Samsung Research America