MetaSym: A Symplectic Meta-learning Framework for Physical Intelligence
Pranav Vaidhyanathan, Aristotelis Papatheodorou, Mark T. Mitchison et al.
Action editor: Yi Liu
https://openreview.net/forum?id=MV1wfMe647
#metasym #symplectic #quadrotor
Long before #multirotor drones became a common sight, the Aerotechnik WGM21 was already pioneering flight as one of the world’s earliest #quadrotor #aircraft. With its ambitious and unconventional design, it explored the concept decades before it became one of the standard technologies in #aviation.
🎉 Thrilled to share our paper accpetance in IEEE RA-L. to be presented at ICRA 2026, Vienna 🇦🇹
🧩 Introducing KQ-LMPC: the fastest open-source data-free Koopman MPC for quadrotors.
🔗 Open-source code: lnkd.in/eCBRzz98
📄 Pre-print (extended): lnkd.in/ey5N9pXz
#Koopman #MPC #Quadrotor #ICRA #Robotics
Real-Time Framework Boosts Quadrotor Connectivity
The framework uses control barrier functions in a unified MPC scheme to keep quadrotors connected and collision‑free; four Crazyflie nano‑quadrotors maintained links and recovered after a loss. Read more: getnews.me/real-time-framework-boos... #quadrotor #crazyflie
RoVerFly Enables Robust Hybrid Control of Quadrotor‑Payload Systems
RoVerFly uses a single reinforcement-learning policy to control quadrotor drones with cable-suspended payloads, handling varying masses and cable lengths without retuning. Read more: getnews.me/roverfly-enables-robust-... #roverfly #quadrotor
JuggleRL: Reinforcement Learning Enables Quadrotor Ball Juggling
JuggleRL lets a quadrotor with a racket juggle a ball, averaging 311 hits over 10 trials and a peak of 462, far above the model‑based baseline’s 3.1‑hit average. Read more: getnews.me/jugglerl-reinforcement-l... #jugglerl #reinforcementlearning #quadrotor
Contextual Neural Moving Horizon Estimation Enhances Quadrotor Control
The new Contextual NeuroMHE system lets quadrotor drones adapt on‑the‑fly and cuts max position error by 20.3% versus prior methods. Read more: getnews.me/contextual-neural-moving... #quadrotor #bayesianoptimization
SDC-Based MPC Boosts Real-Time Quadrotor Control Efficiency
A new SDC‑based model predictive control framework cuts NMPC computation time by over 30 % while keeping high‑precision tracking for fast quadrotor flight. getnews.me/sdc-based-mpc-boosts-rea... #sdcmc #quadrotor #mpc
Multi‑objective PID Controller Optimization for Quieter Quadrotor UAVs
Gradient‑free optimisation of quadrotor PID gains found Grey Wolf Optimization gave best balance of tracking accuracy, power draw and noise, cutting acoustic emissions without hardware changes. getnews.me/multi-objective-pid-cont... #quadrotor #pid
Perception-Aware MPPI Boosts Quadrotor Navigation in Unknown Terrain
PA‑MPPI at 50 Hz cuts navigation time by 100% versus standard MPPI in tests with large obstacles. It serves as a low‑level policy for navigation, turning abstract goals into flight paths. getnews.me/perception-aware-mppi-bo... #quadrotor #mppi
Carnegie Mellon presents LLM-Drone for aerial additive manufacturing Carnegie Mellon University has presented LLM-Drone, a system that combines large language models (LLMs) with drones to expand ad...
#Research #AprilTags #Bitcraze #Carnegie #Mellon […]
[Original post on 3dprintingindustry.com]
Hybrid Neural Process Improves State Estimation of Nonlinear Dynamics
A physics‑informed attentive neural process paired with split conformal prediction delivers 90% confidence intervals for state estimation on a six‑degree‑of‑freedom quadrotor. Read more: getnews.me/hybrid-neural-process-im... #quadrotor #conformal
