Bee's eye view could help robots, drones

Scientists investigating how bees see colour say the insects' highly efficient visual system could revolutionise the way robots and drones view the world.

The way humans see colour is heavily affected by the changing light around them, such as during a sunset or in the middle of the night, but bees see the same colour regardless.

The Melbourne-based team has studied how bees solve this problem, by using three special eyes on top of their head, in addition to two main eyes at the front.

"The three eyes point skyward, and they directly sample the colour of the light above us," Dr Adrian Dyer of RMIT University said.

"If it's a sunny day or a cloudy day, [those three eyes] can detect that.

"It means their brain knows what kind of lighting conditions they are in and then, when they are looking directly at a flower, they can say, 'Ah, it's a blue sky day, so the correct colour should appear like this, or if it's a cloudy day it should appear like something else.'"

This ability is vital to bees, as it allows them to find the best flowers to collect food to take back to their hive.

The team discovered that the three eyes on top of the bee's head, called ocelli, contain two colour receptors that are perfectly tuned for sensing the colour of ambient light.

The information from the ocelli is integrated with the colours seen by the two front eyes.

To prove this was happening, the researchers mapped "neural tracings" sent from the ocelli, showing they feed into the areas of the bee's brain that processes colours.

The project was completed by a joint team from RMIT University, Monash University, University of Melbourne and Deakin University and their findings were being published today in the scientific journal, Proceedings of the National Academy of Sciences of the United States of America.

Technology to copy the bees

Just as humans cannot see colours reliably, because the light around us changes, cameras, robots and drones have the same problem.

Dr Dyer said this long-running conundrum has held back the advance of technology to perform a range of tasks.

"This is a very big problem for machine vision - how to make reliable decisions when the colour of the light changes," he said.

"So our study looked at how bees solve the problem."

The team discovered the mathematical principles used within a bee's brain to process data collected by the three top eyes about the surrounding light and make an adjustment for what the two main eyes are seeing in front of the bee.

The mathematical formula can be easily programmed into a computer.

One example of how the technology could be used is in horticulture.

Drones could fly around an orchid, possibly at night, and accurately detect the colour of different fruits to determine whether they are ready for picking.

The manager can then make accurate decisions about how to deploy his workforce.

This more accurate ability to analyse colour could also be used by drones to inspect infrastructure, such as bridges, or to analyse mineral sands.

"These ideas have been around for a while, but the problem has been how to judge colour accurately," Dr Dyer said.

The project was supported by a grant from the Australian Research Council.

 

Photo: 123RF