International researchers have spent ten years researching the homing instincts of wasps so that they might one day create savvier autonomous robots.
In a study published Thursday in the journal Current Biology, authors describe investigating the methods wasps (Cerceris arenaria) use to find their way back home once they leave their nests. The researchers studied how the creatures moved their heads and bodies, then recreated a simulated "wasp's-eye-view" of the world.
The video—which emulates a wasp's perspective—reveals a distorted and trippy panoramic landscape move from side to side as the wasp moves its head, before it makes beeline for the ground.
"Wasps perform some very specific flight manoeuvres every time they leave home. These are called 'learning flights,'" said Norbert Boeddeker, a biologist at the University of Bielefeld and one of the paper's authors, over the phone. "We wanted to know how this movement helps them return home."
To find out, the researchers recorded a wasp's head movements using high-speed and high resolution stereo cameras. They then decoded this data, and reproduced the wasp's movements by attaching a 360 degree panoramic camera (similar to how a wasp would see the world) to a robot and mimicking how the insect travelled through the air. The researchers discovered that during their "learning flights," a wasp rises from the group and flies from side-to-side in ever-increasing arcs while looking back at its home, in order to get to grips with where it is in relation to its environment.
In the future, the researchers hope to apply their findings to autonomous robots—anything from household Roomba vacuum cleaners to combine harvesters.
"Modern vacuum cleaners go randomly through the room. They don't have a real idea where they are and where they could get power," said Boeddeker. "For a general navigation of the room, it would be nice to know which things are permanent, which things change. This is probably why the wasps do these flights."
Boeddeker said that wasps performed learning flights in order to adapt to different circumstances. He explained that it would be great if future autonomous robots could mimic those techniques by taking snapshots of their environment in order to avoid crashing into unexpected obstacles.
"The wasp's routines and rules for learning might actually help to distinguish the reliable from the unreliable things," he said. "That's also something that maybe a robot could benefit from."