# ROS Aerial Robotics Community Interest Group

## Meeting Notes - April 30th, 2026
Forum Post: https://discourse.openrobotics.org/t/aerial-robotics-meeting-30th-april-2026/54105
**Hosts**
* Mayank Joneja
* Alexis Guijarro
**Attendees**
* Rick Hernandez
* Kimberly McGuire
* Pratik
* Ayham Alharbat
## Agenda
* Announcements
* Events
* Latest developments
* Presentation
* Discussion (if time)
* Announcement next meeting
## Announcements
* Mayank and Alexis are the new leads, and this is our first meeting :the_horns: :partying_face:
## Events (past and future)
### Past-ish
* ROS2 Lyrical Luth testing party earlier today!
https://share.google/ar0hAnqQOFIE0m02e
### Future
* **Xponential 2026**
* May 11-14 2026 Detroit, USA
* https://xponential.org/
* **PX4 Developer Summit 2026 part 1**
* 18-20 May Minneapolis, USA
* Part of [Open Source Summit North America](https://events.linuxfoundation.org/open-source-summit-north-america/)
* Full schedule expected on the PX4 website next week @Ramon?
* **ICRA 2026**
* June 1-5, Vienna Austria
* Workshops:
* [Digital Twins for UAV Autonomy: From Model-Based Design to Verified Deployment ](https://www.mathworks.com/company/events/tradeshows/icra-2026-5136850.html)
* [Aerial inspection for marine infrastructures](https://icra2026.aerialmaritimeinspection.com/)
* [AI-Driven Safe Aerial Robotics](https://aerial-robotics-workshop-icra.com/ )
* [Challenges and Opportunities of Neuromorphic Field Robotics and Automation](https://nfr-icra2026.com)
For more events keep checking the [aerial robotics landscape event page](https://ros-aerial.github.io/aerial_robotic_landscape/events/)
## New projects within the aerial robotics landscape
* One PR for [control systems resources](https://github.com/ROS-Aerial/aerial_robotic_landscape/pull/174)!
* If you'd like to see your project here make a PR!
* Suggestion - Come up with a version of the [Robotics Skill Tree](
https://discourse.openrobotics.org/t/someone-made-a-robotics-skill-tree/38319) :smile:
https://ros-aerial.github.io/aerial_robotic_landscape
## Presentation
"MIGHTY: Real-Time Trajectory Optimization for UAVs Using Hermite Splines"
[Kota Kondo](https://www.linkedin.com/in/kkondo) from [MIT ACL](https://acl.mit.edu/)
https://kotakondo.github.io/
Link to the paper:
https://arxiv.org/html/2511.10822v1
### Notes
1. MIT ACL lab - navigate around potentially dangerous enviornments
multi/single agent navigation
2. Path planning - dynamic + static obstacles
Prev work:
Robust MADER - multiple agents
SANDO - dynamic obstacle avoidance, planning safe trajectories
Main problem - Commercial Solver - Gurobi
Robust MADER - fix time trajectory, commercial solver, slow
MIGHTY - Init trajectory for 5-6 seconds
Spatio-temporal optimization, No commercial solver, fast
Quintic (5th degree) Hermite Splines
More degrees - easier to deform, harder to solve
Hermite spline - P0 starting, P3 - goal point, get P1, P2 - waypoints, velocity and acceleration vectors at waypoints
Can even control how much time you spend in each state
Hermite Splines Features:
- Full polynomial + time space
- C2 continuous cost
- Quasi Newton method for non-convex optimization
Outperform EGO-Swarm V2 and SUPER in flight simulation for static obstacles
Navigation operations are solved in real-time, already tested on real hardware with onboard computers.
Current efforts are aimed to adapt the approach into multi-agent systems.
The code is open sourced and published in Github: https://github.com/mit-acl/mighty
## Discussion
* Would you be interested in getting pre-reads as a part of the announcement? e.g. Kota's paper? - Community said YES! :smile:
* (Pratik Phadte) What's the onboard computing platform that you have used. A: Intel NUC, but we have been tested the solution on the Jetson Platform.
* Can I use a stereo camera? A: You can use any stereo or depth camera, meanwhile you can get a map representation through depth information.
* (Ayham Alharbat) Does Mighty take the dynamics of the robot and its constraints into account? A: We used polynomial representation and dimensional flatness, if the trajectory is contrained into the right manner it can be mapped to be used by MIGHTY.
* (Rick Hernandez) Is Multi-Mighty information available for review? A: The paper is not officially published yet, you can review the github code.
* (Chethan Danivas) How do you handle tracking into MIGHTY? A: MIGHTY assumes some error margin so the drone can go through, which is half of the size plus some margin, how to handle the error is an unexplored territory.
* (Naman Raj) Where to start for simulating drone projects, path planning, trajectory prediction, etc and all? A: https://ros-aerial.github.io/aerial_robotic_landscape/simulation/ , https://ros-aerial.github.io/aerial_robotic_landscape/education_and_tutorial/
* (Mayank) Commercial solvers and Abstraction layers to be used on hardware
* (Chethan Danivas) Why to use the solver onboard? A: Because it can present a more robust solution, although if you can use onground solutions.
* (Mayank) Are you planning to use GPU, like the one embedded on the Jetson platform? A: This is an option as it presents some benefits aside of performance but also mathematical, like to avoid local minimals.
## Next meeting
On the 28th of May! Presentation and topic still to be determined.
Add the [ROS community event calendar ](https://calendar.google.com/calendar/u/0/embed?src=c_3fc5c4d6ece9d80d49f136c1dcd54d7f44e1acefdbe87228c92ff268e85e2ea0@group.calendar.google.com&ctz=UTC)to your calendar to see our planned meetings.
We also post an announcement [Aerial Vehicles category](https://discourse.ros.org/c/aerial-vehicles/14) on ROS discourse with the [wg-aerial-robotics tag](https://discourse.ros.org/tag/wg-aerial-robotics) about one week before the meeting starts.
<iframe src="https://calendar.google.com/calendar/u/0/embed?src=c_3fc5c4d6ece9d80d49f136c1dcd54d7f44e1acefdbe87228c92ff268e85e2ea0@group.calendar.google.com&ctz=UTC" style="border: 0" width="800" height="600" frameborder="0" scrolling="no"></iframe>
### Want to nominate someone/yourself for a presentation?
Send us a message on the #aerial-robotics channel channel (link below) we are always looking for cool open source projects to spotlight.
**👉 Reach out to Mayank / Alexis.**
### Handy Links
* [ROS Discourse wg-aerial](https://discourse.ros.org/tag/wg-aerial-robotics)
* [ROS Aerial Community Github](https://github.com/ROS-Aerial)
* [#aerial-robotics Zulip Channel (Open Robotics Server)](https://openrobotics.zulipchat.com/#narrow/channel/526057-Aerial-Robotics)
* [Aerial Robotic Landscape](https://ros-aerial.github.io/aerial_robotic_landscape/)