Getting a humanoid robot to jump is no easy task <... You need to have very powerful servos capable of very high acceleration, something that until recently wasn't readily available to hobbyists. Not in a small package and relatively reasonable price, anyway. This robot is able to jump some 3cm into the air using the new (as of August 2009) Kondo KRS-2552HV in its legs; these servos put out 14kg*cm of torque and can rotate 60 degrees in 0.14 second. They're no RX-64, but they only cost about $80 each as opposed to $300.

The upshot of all this is that servos are getting faster, stronger, and cheaper, which means that we’ll be seeing more robots capable of feats like running and jumping.

VIA [ Robots-Dreams ]

Remember that jumping grasshopper robot from May of last year? It still hasn’t quite figured out how to fly, but it can now make more than one autonomous jump in a row, thanks to a primitive simple but effective self-righting system. It’s the same type of thing we saw on the WeebleCopter: a spherical metal framework with the robot on the bottom, where its weight will cause the whole thing to roll upright:

There are some downsides to this system, including increased bulk and most notably a decrease in jumping height of nearly 25%, but the frame does protect the robot, and if it gets stuck in a tight spot, it can use the frame to bounce off obstacles and get itself pointed in a different direction.

[ Self Deploying Microglider @ EPFL LIS ]

When we posted about the Precision Urban Hopper last week, we commented that one of the biggest problems with hopping robots is keeping them stable. This WeebleCopter (I’m calling it that, it’s actually called a “hopping rotochute”) neatly solves this problem by using coaxial rotor blades to both jump up and decelerate back down to the ground. A weighted base and spherical roll cage keeps the robot stable in flight and makes sure that it always lands right-side up, Weeble-style. It’s important to note that this robot is not designed to be a helicopter… It’s not, as far as I can tell, capable of fully controllable flight (although it can hover), but instead uses a simple movable mass to control its pitch and direction of travel. This may seem like a disadvantage, but it saves on complexity and weight, and makes the bot more efficient: it can jump over obstacles or up to a perch when it needs to, but the rest of the time, it’s just sitting there and not expending energy.

There aren’t many details on the future of this robot, a project by Eric Beyer and Mark Costello from the Georgia Institute of Technology, but it (shockingly) looks to have been funded primarily by the army. Looks like hopping robots are going to be the next big battlefield toy, we’ll just have to see which design becomes the first to jump across the finish line into production.

VIA [ New Scientist ]

We first heard about the Precision Urban Hopper back in May, as a jumping robot collaboration between Sandia Labs and Boston Dynamics. This video is the first time we’ve seen the thing in action, though, and it seems to be pretty well on track to fulfilling the design concept. I don’t think it would be too much of a stretch to say that the easy (easier) part is done; that is, creating a robot capable of launching itself 25 feet in the air in a specific trajectory with a piston. The thing that doesn’t appear to have been figured out yet is the ability to stabilize the jump. Certainly, the robot works fine as is: it’s durable enough to jump, crash, and keep on driving. But it would be very handy to have a robot capable of making stable jumps, which would allow it to return useful in-flight video… For example, if you want to see what’s over a wall, send this little guy over and have him jump up and take a peek. There are already solutions for this exact problem, but the Precision Urban Hopper offers more flexibility, since the hopping package can be used for both surveillance and movement.

Delivery is still scheduled for late 2010.

[ Sandia Labs ]

This may not look like the most promising design for a robot, but there’s a lot of potential to be had with robots that can change their shape. These robots, from Ritsumeikan University in Japan, are constructed with spherical shells of spring steel attached to an inner core (which contains the power source and electronics) via shape memory alloy wires. Applying voltage to the wires causes them to contract, deforming the shape of the robot. By doing this, the robot can change its center of gravity to roll in any direction, and by contracting the spring steel enough, the robot stores up enough energy to jump.

The big limitation at the moment is that shape memory alloys only really work in one direction: applying a voltage heats the wire, causing it to shrink, but in order for the wire to unshrink, it has to radiate that same amount of heat, which takes a little while… So the robot can deform to roll or jump, but after it does, it needs a cool-down period.

Even with this (surely surmountable) limitation, jumping robots have an advantage over both ground and flying robots: they can move around without respect to terrain without having to expend energy staying airborn. ‘Course, they’re not as efficient as a ground robot, or as versatile as a flying robot, but they’d be perfect for things like planetary exploration where reliability and versatile mobility are more important than speed.

More info is available in this paper (*.PDF).

[ Ritsumeikan University ]

Last May, we wrote about a 7 gram robot grasshopper that is capable of jumping a distance of 1.4 meters, which is pretty huge for such a small robot. By using reduction gears and legs that act as springs, the robot is a very efficient mover, as well. We commented in that post that “the great thing about jumping is that it combines the advantages of being on the ground with one of the most important advantages of being able to fly: obstacle avoidance.” Of course, the other big advantage of being able to fly is that you can cover large distances quickly and efficiently (albeit mostly due to the aforementioned avoidance of obstacles).

Researchers at EPFL’s Laboratory of Intelligent Systems have made their robot grasshopper into a true flier by adding another bit of real grasshopper anatomy: wings. Made of carbon, mylar, and shape memory muscles, the wings are fully controllable and the robot is able to steer itself towards sources of light. Eventually, the wings will be retractable and extendable just like a real grasshopper, and when integrated with the jumping legs and possibly some solar panels, you’ve got a tiny, cheap, extremely efficient robot capable of covering large distances over just about any sort of terrain by jumping, gliding, landing, and repeating.

Currently the wings and body aren’t really able to work together, so pending that, we’ve got videos of the two halves after the jump. (Read more…)