Proceedings · Session S-915 · filed November 15, 2026
AI & Emerging Tech in R&DSession paper
Floaty robot rides updrafts with adjustable wings, holds position at 10 m/s
Floaty, a Max Planck and Stuttgart flying robot, stabilizes itself on updrafts using four adjustable wing flaps, holding position at wind tunnel speeds up to 10 m/s without powered hovering.
By Sophie Lindqvist3 min read677 words
Summary
- Floaty remained stable in vertical wind tunnel airflow up to 10 meters (33 ft) per second.
- Researchers from the Max Planck Institute for Intelligent Systems and the University of Stuttgart developed the robot, published in npj Robotics.
- The robot does not flap its wings; it adjusts four wing flaps to control airflow and air resistance for passive stabilization.
- An early flat prototype flipped over in tests until the team lowered its center of gravity and added a bend to the flaps.
- Prof. Michael Mühlebach cited factory smokestack inspection as the nearest-term application, with rocket re-entry control and weather balloon guidance as possibilities.
A palm-scale flying robot called Floaty has demonstrated passive stabilization in vertical wind tunnel airflow of up to 10 meters (33 ft) per second, using four adjustable wing flaps rather than powered rotors or flapping flight to hold its position. The system, developed at Germany's Max Planck Institute for Intelligent Systems and the University of Stuttgart, appeared in the journal npj Robotics.
The design takes its cue from kestrels and other birds of prey that hover or soar on rising air columns while expending little energy. Floaty does not flap its wings to generate lift. Instead, an operator manually places the robot into an updraft, and the device then adjusts its four movable wing flaps automatically to manage how air flows around the body. By changing its own air resistance, the robot stabilizes itself and can translate both vertically and horizontally within the air column.
That control authority extends to disturbance rejection. Floaty can hold itself in place when gusts push it from the side — a capability the researchers demonstrated, and in one iteration hard-won.
What went wrong in the wind tunnel?
An early, flat version of Floaty flipped over during vertical wind tunnel tests when subjected to lateral disturbances. The team's fix was mechanical rather than algorithmic: they lowered the robot's center of gravity and added a bend to the wing flaps, which had previously been rigid. The revised configuration remained stable at tunnel air speeds up to 10 meters per second, the figure reported with the paper.
For R&D managers weighing energy budgets in aerial sensing or inspection platforms, the underlying proposition is simple: hover without continuously running motors, and the power draw associated with station-keeping largely disappears. The trade-off is equally concrete — Floaty cannot self-deploy. It must be placed into an updraft by hand, which confines near-term applications to environments where rising air is already present and persistent.
Where could it actually work?
The team's own assessment of use cases is measured rather than promotional. "Floaty could inspect factory smokestacks where there is strong upward airflow. It could potentially work there with little modification," said Prof. Michael Mühlebach of the Max Planck Institute. "Similar technology could perhaps also help control rockets during re-entry, or it could help guide weather balloons. There are many ways in which the robot can take advantage of upward airflows to save energy."
The smokestack scenario is the most immediate: a structure that generates its own updraft, where a robot that rides rather than fights the airflow could linger over the stack for monitoring tasks without the battery burden of a quadcopter. The rocket re-entry and weather balloon applications remain speculative in Mühlebach's framing — both are flagged with "perhaps" and "could," and neither carries test data in the published results.
What is measured versus projected?
The measured result in the source material is the stability demonstration: sustained position-holding, including under lateral gusts, at wind tunnel speeds up to 10 m/s, achieved through passive aerodynamic control via wing-flap adjustment. The deployment concepts — stack inspection, re-entry control, balloon guidance — are projections from the authors, not validated field trials, and the source does not report flight endurance figures, payload capacity, outdoor tests, or the robot's mass and dimensions.
The work also leaves open questions of autonomy beyond the air column. Floaty controls itself once inside an updraft, but the current workflow requires manual placement, meaning any operational system would need a delivery mechanism — a launch vehicle, a second drone, or a fixed mounting — before the energy-saving benefits apply.
The research is published in npj Robotics, and the Max Planck–Stuttgart group positions the energy argument as the core claim: upward airflows are free, and a robot that exploits them for lift and stability avoids the battery drain that limits conventional hovering drones.
Having proven gust-resistant stability at 10 m/s in controlled tunnel conditions, Floaty now faces the step from laboratory airflow to the messier, shifting updrafts of real stacks and skies.
via newatlas.com (Source)
Filed under
- robotics
- updrafts
- passive-flight-control
- wind-tunnel-testing
- adjustable-wing-flaps
More from Sophie Lindqvist
Show full bio
Correspondent covering money, funding and market shifts around research and development at Hypothesis Wire.
242 articles
References
- Hopcopter foot pad cuts cross-terrain performance variation by 84%
- Sandia tests reinforcement learning on cooperative drone evasion
- 3D-Printed Robot Frog Swims 80 mm on Light-Powered Artificial Muscles
- Kyushu team reports self-docking flexible circuits with sub-100 μm actuators
- Skyways claims 1,300-mile cargo drone at a tenth of helicopter cost