Stanford University researchers have developed a robot that can catch fruit flies without injuring them – an invention promising to open new avenues in biomedical research.
The robot, resembling a miniature flying saucer, uses a narrow suction tube to lift up a fly selected from a bunch and inspects it with its machine vision capability. It can either analyse it without injuring it or, alternatively, perform micro-dissections revealing the insect’s tiny organs.
"Robotic technology offers a new prospect for automated experiments and enables fly researchers to do several things they couldn't do previously," said Mark Schnitzer, an associate professor of biology and applied physics and creator of the robot.
"For example, it can do studies with large numbers of flies inspected in very precise ways."
Using the robot, the researchers were able to study 1,000 flies in 10 hours, a task that would have taken much longer to even a highly skilled human lab worker. As fruit flies are the number one experimental object in biomedical research, the device could substantially speed up developments in genetics and biomedicine.
Moreover, the researchers believe, the apparatus could for the first time enable performing behavioural studies on fruit flies, which was previously impossible.
In the study, published in the Nature Methods journal, the team described how they used the robot to pick up a fly and expose it to different smells while recording its reactions.
Despite fruit flies' omnipresence in biomedical laboratories around the world, researchers have so far struggled to create an automated machine to study the insects – mostly because of their speed. Instead, the flies had to be drugged to stay manageable.
The Stanford University robot can recognize each individual fly by its reflection pattern, analyse its physical attributes and differentiate males from females.
In one experiment, the robot's machine vision was able to differentiate between two strains of flies so similar they are indistinguishable to the human eye.
"At the beginning we were thinking it wouldn't work," said Joan Savall, a senior scientist at the Howard Hughes Medical Institute, who participated in the study. "It's not just picking the flies up, it's keeping them alive.”
In addition to not squashing the flies, the robot needed to be able to distinguish between the flies it had observed to prevent analysing the same fly repeatedly. "The key was flashing an infrared light," Savall said. Not visible to the flies, the light does not disrupt their behaviour but provides a visual pattern the robot could use to distinguish between individuals.
Savall says freeing graduate students to do science rather than toiling in the fly room is one benefit of the robot, but even more important is the range of new experiments that will be possible. "In the end you can really push many fields at the same time," he said.