‘FlyingToolbox’ drone system achieves accurate mid-air tool exchange despite airflow interference

Flying manipulator robots have shown themselves to be useful in many applications, such as industrial maintenance or construction. Their utility in hard to reach or hazardous locations makes them particularly promising in applications that put humans at risk. While these machines have been continuously improving over the years, they are still lacking in certain areas.

One difficulty for drones in the past has been the ability to stack on top of one another and work cooperatively while in flight. This ability is useful for things like swapping tools, similar to the way a nurse might hand different tools to a doctor during a procedure—allowing the doctor (or manipulator drone) to work uninterrupted.

The difficulty comes from something called “downwash,” which is a strong movement of air generated between two drones that interferes with their precise movements and docking procedures. However, a team of researchers from Westlake University in China has designed a new system of micro-aerial vehicles (MAVs) capable of exchanging tools with impressive precision while flying. The design and experimental tests on the “FlyingToolbox” are documented in their new study, published in Nature.

The FlyingToolbox consists of an MAV that holds multiple tools, along with a robotic arm manipulator MAV. The system uses onboard vision (QR code tracking) for precise relative positioning, a neural network-based estimator to predict and compensate for downwash disturbances in real time, and an electromagnet docking mechanism with elastic tethers for tool attachment.

“During the docking process, the propellers of the manipulator MAV generate persistent and intense downwash disturbances on the toolbox MAV underneath. For instance, when the vertical distance between the two MAVs is 0.6 m, the downwash speed can reach 13.18 m/s, exerting an unsettled disturbance up to 24.9 N, which is 40.2% of the weight of the toolbox MAV (6.32 kg). To counter the downwash airflow, it is important to estimate and then compensate for the induced disturbance force and torque,” the study authors explain.

To test whether the system would be able to compensate for the force of the downwash, the team conducted multi-stage tasks and tool switching experiments with both stationary and moving toolbox drones, and then conducted 20 consecutive docking trials to test for accuracy and repeatability.

Not only did the multi-stage tasks successfully prove the system was capable of tool-switching in flight, but the FlyingToolbox also achieved a sub-centimeter docking accuracy of 0.80 ± 0.33 cm, even when a strong downwash of up to 13.18 m/s was present. The team says this is a major improvement when compared to the accuracy of similar systems, which achieved accuracies between 6–8 cm.

The FlyingToolbox appears to be a big step forward in the world of flying manipulator robots, with many potential applications, but there are still a few hurdles to clear. For example, the drone system was tested out in a controlled lab, so functioning in outdoor conditions may prove more difficult.

However, the team is pleased with the progress. They say, “The accuracy, robustness and versatility of the system make it a potential solution for a wide range of real-world tasks. We believe that it may inspire the community to develop complex cooperative aerial manipulation systems, for example, for battery replacement and material replenishment, substantially expanding the abilities of aerial manipulators.”

© 2025 Science X Network