Back in the 80s, specifically in 1989, the publication of an article entitled Fast, Cheap, and Out of Control: A Robots Invasion of the Solar System. Its authors, Rod Brooks and Anita Flynn, proposed in this text the idea that instead of sending a large robot to explore the surface of Mars they could send a large number of robots small, simple and extremely cheap To achieve this goal. At the end of that work the authors conclude the following:
We suggest that within a few years it will be possible, at modest cost, to invade a planet with millions of tiny robots.
Currently we are still far from being able to send millions of tiny robots to another planet. And it is that building robots the size of an insect constitutes a process that requires significant progress that make their autonomy possible.
However, progress is already being made within this project, which was published in an article by Washington University adjunct professor, Sawyer Fuller where he explains the process carried out to develop the detection hardware that will be implemented in flying robots, which will have smaller than a grain of rice.
In this sense, the autonomy of a tiny flying robot will be determined by its ability to maintain its own stability, as well as by its ability to know where you are located and travel where you want.
The latter can be somewhat complex to carry out for small flying robots, due to their susceptibility to air currents or turbulence that may interfere with the flight process towards its destination.
However, there are some advantages present in this type of robots, among them the fact that since the air resistance is more dominant, the gyroscope implanted in the robot does not offer any utility, but it does an accelerometer which can be joined to an optical flow chamber to follow the movement and a microcontroller to perform the calculations.
In the end, Fuller achieved a hardware that weighs 6.2mg and consumes energy 167 microwatts. All this in theory should be enough to make a robot of 10 mg (the size of a mosquito) take flight
However, since robotic mosquitoes don’t exist, checking whether or not this hardware works is difficult.
This prompted the researchers in charge of the project to resort to using a palm-sized drone and equipping it with simulated sensors.
At the time of carrying out the tests, it was possible to demonstrate how the system was able to successfully estimate the attitude of the droneas well as detect and rule out disturbances generated by the wind.
In that sense, the drone showed a performance comparable to that of a real fruit fly, certainly a remarkable result considering the little time it took to perfect its design.
That’s how Fuller feels confident considering that the mosquito-sized robots will give him a whole new approach to autonomous exploration. In that sense, he added:
Small flying insects will revolutionize low-altitude atmospheric telemetry, that is, the remote sensing of airflow and composition, and will do so in much more detailed and persistent ways than is currently possible.
