Paper plane-throwing robot provides insights into aerodynamics for engineers

Paper planes aren't just a child's toy; we can learn a lot about aerial vehicle design by modelling & optimizing their unique & unpredictable flight behavior. And that's what @EPFL_CREATE has done in their latest @SciReports paper. Congrats! @EPFL_MechEhttps://t.co/v7nzfUuEuf pic.twitter.com/SdEGOfBBbD

— EPFL School of Engineering (@EPFLEngineering) March 14, 2023

The robot designs simple paper planes, flies them, and analyzes data based on the planes’ performance, all without any human interference. The team made the robot arm construct and fly 500 paper planes to observe the true probabilistic and stochastic nature of the flight behavior. 

Once the plane was up in the air, ​​the 2D trajectory of each flight was recorded via a camera. All 500 trajectories were recorded. Once the team observed all 500 flights, they found that their trajectories could be divided into three behavioral groups. 

The first was the nose dive, meaning the paper airplanes had a short flight distance and hit the ground as soon as they escaped the clutches of the robot arm. The second was glide, meaning that the paper planes had a longer flight distance than those that nose-dived and glided in the air for a bit. The third behavior group was recovery glide, in which the paper planes had the longest flight distance, often with a small upward trajectory in the middle of the flight.

The researchers observed that ‘the airplane flight trajectory behaviors are complex, where the mapping between the geometry and behavior is unintuitive, probabilistic in nature, and cannot be solved analytically.’ This means that even if the paper was folded the same way each time, it’s not guaranteed to fly the same distance and follow a similar trajectory.

“Exploiting the precise and automated nature of the robotic setup, large-scale experiments can be performed to enable design optimization. By testing and evaluating many airplanes, the design space can be characterized and explored… we demonstrate how developing these models can be used to accelerate real-world robotic optimization of a design—to identify wing shapes that fly a given distance," researchers noted.

The study was published in Nature on March 14.

Study abstract:

Although often regarded a childhood toy, the design of paper airplanes is subtly complex. The design space and mapping from geometry to distance flown is highly nonlinear and probabilistic where a single airplane design exhibits a multitude of trajectory forms and flight distances. This makes optimization and understanding of their behavior challenging for humans. By understanding the behavior of paper airplanes and predicting flight behavior, there is a potential to improve the design of aerial vehicles that operate at low Reynolds numbers. By developing a robotic system that can fabricate, test, analyze, and model the flight behavior in an unsupervised fashion, a wide design space can be reliably characterized. We find there are discrete behavioral groups that result in different trajectories: nose dive, glide, and recovery glide. Informed by this characterization we propose a method of using Gaussian mixture models to extract the clusters of the design space that map to these different behaviors. This allows us to solve both the forward and reverse design problem for paper airplanes, and also to perform efficient optimization of the geometry for a given target flight distance.