Last week, we wrote about a little robot that can be steered in any direction, even though it only has one motor. This week, we thought we’d follow it up with some flying robots that are also taking advantage of clever design to find the absolute minimum number of motors required for stable flight. And as you may have guessed from the headline, it’s not very many motors at all.
We’re used to seeing two types of hovering robots. The first type is the more traditional helicopter design, with a single main rotor (or two) and maybe a little tail rotor. The second type is the pervasive quadcopter (or hexacopter or octocopter or whatever). With a helicopter, you have to manage a very complex multiple-actuator likage system for control and stability, and quadcopters have at least as many motors as they do rotors.
The Modular Robotics Laboratory at the University of Pennsylvania has been seeing what it takes to reduce both the complexity and number of actuators of rotorcraft, and they’ve come up with some vey cool solutions. Let's start with seeing what the minimum number of actuators that you need for controllable flight is:
To clarify what’s going on here: that rotor (the one at the top; the bottom one is fixed) has to alter its angle of attack every single time it spins around, which is 40 times every second. It’s a complex thing to control, and if you’re trying to do it with just one motor, you have to somehow manage to make it work with the only thing you do have control over: the acceleration of said motor (which is how 1STAR works too).
In the case of the robot in the video, the clever bit is a passive linkage that translates motor acceleration into pitch changes. By pulsing the motor (adjusting the magnitude and response time) when the rotor blades are at just the right orientations, their pitch can be adjusted fast enough, and with a fine enough degree of control, you get full authority over pitch and roll.
You’ll notice that with the MAV in the above video, there’s a second motor driving a second rotor to cancel out the angular momentum from the first rotor, as well as to control yaw. If you get rid of that motor, besides losing yaw control you’ll also end up with a robot that’s spinning around really fast all the time.
Quadcopters don’t have this problem, because they have an even number of spinning rotors. Conventional helicopters use a tail rotor to counteract it. But if you’re looking to simplify, are there any other options for creating a MAV that’s passively stable, meaning that it can hover in an arbitrary position in space without any control inputs? Why yes, there are:
And thanks to UPenn for being awesome and showing us what happens when their robots don’t work.
By reducing the number of motors and actuators, the price of the robot drops drastically, and its weight goes down and endurance goes up, making it much more versatile. And as with 1STAR, if you want to make a lot of robots, low cost and simplicity become some of the most important design features.
The obvious question that we have now is, is there any reason why you couldn’t combine these two platforms, yielding a single motor, single rotor flying robot that’s both passively stable and fully controllable in pitch and roll? Maybe there is a reason, but if there isn’t, you can bet we’ll see a prototype from UPenn at a robotics conference sometime soon.
“Passive Stability of a Single Actuator Micro Aerial Vehicle,” by M. Piccoli and M. Yim, was presented at ICRA 2014 in Hong Kong. “An Underactuated Propeller for Attitude Control in Micro Air Vehicles,” by J. Paulos and M. Yim, was presented at IROS 2013 in Japan.
[ MODLAB: Smart, Underactuated ]
Evan Ackerman is a senior editor at IEEE Spectrum. Since 2007, he has written over 6,000 articles on robotics and technology. He has a degree in Martian geology and is excellent at playing bagpipes.