news
[DSK Conference] Physical AI, why is it "drone"
2026-02-25

"The answer to physical AI is drones."

This was emphasized at the 'DSK 2026 Conference' held in BEXCO, Busan, for three days from the 25th. "Drones are winged robots," said Jeongwan Koh, co-founder and director of operations at WEFLO.

Until now, the product represented by physical AI has been humanoid robots. But why do field experts point to drones as physical AI products.


◇ Instead of humanoids, drones expand

Humanoid robots take over what humans used to do. It looks like a human being and can use the tools that people used. Eventually, it moves in a place where people live and work, that is, on the ground. Drones are different. It travels through spaces where humans cannot step. Fire scenes where flames are soaring, battlefields where GPS and communication are cut off, and outer walls of buildings that are inaccessible to people. The drone makes its own judgment and completes its mission there.

"If humanoid robots replace what humans were doing, drones solve problems in three-dimensional space that humans could not do," said Jeongwan Koh, COO of WEFLO. 


COO Jeongwan Koh of WEFLO, speaking at DSK 2026 Conference ( Image=Reporter Dongwon Kim)


◇ Robots that mount physical AI into the brain, what are the walls that need to be crossed?

It is clear that drones are a stage for the possibility of physical AI, but technical challenges remain. COO Koh said, "There is also an idea that it is honestly not easy to mount physical AI on drones." Because drones have to fly in the sky, they are so restricted that they cannot be compared to robots on the ground.

The more sophisticated AI models are, the more computations are required, and the more computations are, the greater the battery consumption. Currently, the flying time of drones is only one to two hours, and batteries account for 30-50% of the total weight. The more AI tries to improve its performance, the heavier the drone becomes and the shorter the flight time is, the more the vicious cycle repeats.

COO Koh pointed out, "Just as smart glasses have been stuck in place for a long time due to hardware limitations, drone physical AI can also hit the same wall if the battery and weight reduction problems are not solved."









TAG
DSK
Physical AI
AI
Conference
Drone
LIST
TOP