Have you ever wished for an automaton that can get the party started, raise the roof, and all that? You’ll want to meet [DJ Pfeif]’s Flippin Rhobot, then. He’s a hype bot from the world of Twitch streaming, and he apparently knows how to party.
Flippin Rhobot is controlled by an ESP32 that listens into the chat on [DJ Pfeif]’s stream. He’s got a vaguely humanoid form, and he can rotate on the spot and wave his arms in the air courtesy of a few servos. He’s also got a little computer terminal that displays the show’s “Hack the Planet” logo when he turns to face the screen. His body also features some addressable LEDs that flash and dance on command.
[DJ Pfeif] does a good job of explaining the project, and includes the code that laces everything together. Interfacing with Twitch chat can be fun, and we’ve featured a guide on doing just that before, too.
Human hands are remarkable pieces of machinery, so it’s no wonder many robots are designed after their creators. The amount of computation required to properly attenuate the grip strength and position of a hand is no joke though, so what if you took a tentacular approach to grabbing things instead?
Inspired by ocean creatures, researchers found that by using a set of pneumatically-controlled tentacles, they could grasp irregular objects reliably and gently without having to faff about with machine learning or oodles of sensors. The tentacles can wrap around the object itself or intertwine with each other to encase parts of an object in its gentle grasp.
The basic component of the device is 12 sections “slender elastomeric filament” which dangle at gauge pressure, but begin to curl as pressure is applied up to 172 kPa. All of the 300 mm long segments run on the same pressure source and are the same size, but adding multiple sized filaments or pressure sources might be useful for certain applications.
Imagine you want to iterate on a shock-absorbing structure design in plastic. You might design something in CAD, print it, then test it on a rig. You’ll then note down your measurements, and repeat the process again. But what if a robot could do all that instead, and do it for years on end? That’s precisely what’s been going on at Boston University.
Inside the College of Engineering, a robotic system has been working to optimize a shape to better absorb energy. The system first 3D prints a shape, and stores a record of its shape and size. The shape is then crushed with a small press while the system measures how much energy it took to compress. The crushed object is then discarded, and the robot iterates a new design and starts again.
The experiment has been going on for three years continuously at this point. The MAMA BEAR robot has tested over 25,000 3D prints, which now fill dozens of boxes. It’s not frivolous, either. According to engineer Keith Brown, the former record for a energy-absorbing structure was 71% efficiency. The robot developed a structure with 75% efficiency in January 2023, according to his research paper.
The former record holder was measured at 141 mm, so [Lidor] had to beat that. He set about building a humanoid robot 95 mm tall, relying on off-the-shelf parts and 3D-printed components of his own design. An ESP32 served as the brains of the operation, while the robot, named Tiny Titan, got big flat feet to make walking relatively stable and controlled. Small servos were stacked up to actuate the legs and create a suitably humanoid robot to claim the title.
Sadly, [Lidor] was pipped to the post. Some procrastinating in finishing the robot and documentation saw another rival with a 60mm robot take the record. It’s not 100% clear what Guinness requires for someone to take this record, but it seems to involve a robot with arms, legs, and some ability to walk.
While it doesn’t look like a traditional robot, the hydrogel robot from [Zi Liang Wu] forms a möbius strip and can be activated by light. They also experimented with shaping the hydrogels as a Seifert ribbon.
The key is that the hydrogels contain gold nanoparticles. Light heats the gold particles and this causes the hydrogels to move. The connections between the strips of hydrogels causes them to move in predictable ways. You can see a video about the experiments below.
These robots aren’t going to be for warehouse or factory work. But they can do tasks like collecting plastic beads, something difficult for conventional robots to do. They also hope to demonstrate that these soft robots could work in the body for taking samples or delivering a drug, although it isn’t apparent how light would get to them inside your body.
The dark side of the material tends to turn towards the light. The continuous loop structure means it never runs to the end of its travel. Watching it move on a string is pretty impressive.
Crawling and slithering robots may be the answer for certain specialized applications. After all, it works well in nature.
Most of the Hackaday community would never wire a power supply to a circuit without knowing the expected voltage and the required current. But our mechanical design is often more bodged. We meet folks who carefully budget power to their microcontroller, sensors, and so on, but never measure the forces involved in their mechanical designs. Then they’re surprised when the motor they chose isn’t big enough for the weight of their robot.
An obstacle to being more numbers oriented is lack of basic data about the system. So, here are some simple tools for measuring dynamic properties of small mechanisms; distances, forces, velocities, accelerations, torques, and other things you haven’t thought about since college physics. If you don’t have these in your toolkit, how do you measure?
Distance
For longer distances the usual homeowner’s tools work fine. The mechatronics tinkerer benefits from two tools on the small end. A dial or electronic caliper for measuring small things, and a thickness gauge (or leaf gauge) for measuring small slots.
A thickness gauge is just metal leaves in different thicknesses, bolted together at one end. Find a combination of leaves that just fits in the space.
Force
Here’s four force measuring tools we use to cover different magnitudes of force: a postage scale, a push stick, a spring scale, and a letter scale. The postage scale is best purchased. For big things, the bathroom scale works.
A push stick is a force measurement device that you can make yourself. We first saw one of these used to tune slot cars, but they’re universally useful. It’s a simple pen shaped device made with a barrel from any small transparent parts tube, a spring, and a plunger with a protruding pin. Grasp the barrel and push the gizmo with the pin, and you can read the force off the tube.
If you need it to be calibrated, remember that you just bought a postage scale. Push it into the scale and mark off reasonable increments. Make several, in different sizes. A Z or L shaped plunger is useful for hard to reach places.
The conventional spring tension scale is useful, but most commercial ones are terribly made and inaccurate. You can make yourself a better one. They are useful for measuring the spring constant of springs, for learning the tractive effort needed to move a robot, finding the center of gravity of a robot arm, and a hundred other ‘how much oomph’ things. Again, it’s just a matter of connecting a hook to a spring, and measuring its deflection.
For yet lighter weights, you could buy a letter scale, at least in the old days. Today you might have to make your own. It can be as simple as a piece of spring steel fixed to a sheet of calibrated cardboard.
Torque
Torque measurements are good not only for sizing actuators, but for measuring efficiency.
How you do torque measurements depend on the speed you want to make them at. For static loads, just put a lever of known length on the shaft and measure the force. Torque = distance * force. For fast rotating systems, you can run the system at a known speed and measure the electrical energy used.
Schematic of a Prony brake by [MatthiasDD]If you just want to apply a varying known torque to measure efficiency, your life is much easier. Mount a broad wheel of some sort on the shaft — RC airplane tires work well. Drape a piece of ribbon over the tire. Anchor it at the “out” end and hang a small weight at the “in” end. This is a Prony Brake, and it’s a useful device to know about. The force on the outside of the wheel is just enough to lift the weight – after that the ribbon slips. The measured torque is then the weight times the wheel radius.
You may also want to measure speeds and accelerations. Here, the ubiquity of cell phone cameras is your friend. Suppose you’re animating a crane on your model railroad. Record yourself on video moving the crane with your hands against a protractor to get a feel for speed and acceleration. In video editing software check the positions for various frames, and you now have position changes. The number of frames and distance can help you calculate the speed, and the change in speed vs time is acceleration.
If your mechanism is moving too fast for video, use a fast phototransistor or hall effect device and an oscilloscope, or gear down by holding a toy wheel against the shaft and measure the more slowly rotating wheel.
In the crane example, the torque you need to supply is the frictional torque plus the acceleration torque, and to calculate the acceleration torque you need the moment of inertia. For refresher: angular acceleration = torque / moment of inertia (ω = τ / I) and moment of inertia = mass * radius2 (I = m * r2 ) for point mass.
You can drive the crane with a repeatable torque, say using a pulley and weight or a motor, and get the acceleration ω1 from the still frames on your video. If you repeat this with a known mass m a known distance r from the shaft axis, like a lump of putty on the end of the crane arm, you can get a second value: ω2.
Write the ω = τ / I equations, ω1 = τ / Icrane and ω2 = τ / (Icrane + r 2 * m). Combining and isolating Icrane and holding our tongues just right, Icrane = r2 * m / (ω1 / ω2 – 1).
Be careful to subtract the moment of inertia of your measuring apparatus, and add in the moment of inertia of the final drive if needed. Now you can size your servo with some confidence. Believe me, once you’ve done this a couple times, you’ll never go back to winging it.
Power
The easiest way to get a ballpark feel for power is to simply measure the system’s consumed power by measuring the electrical power at the motor, but this ignores losses in the drive train. And losses are one of the really interesting things to measure. Bad performance is usually friction, and efficiency is a goal for other reasons than just motor sizing or battery life. It’s a measure of how janky your setup is.
Does your model train or robot run poorly? Set it to climb a steep grade on a test track. Calculate the work it does: mass * height change. Measure the input electrical power and the time, Energy = V*I*T. You now have an idea of how much the actual power consumption differs from the maximally efficient system. Any power that went in but didn’t appear as potential energy in the choo-choo’s new position is frictional loss. Now you can experiment with loosening and tightening screws, changing gear mesh, and such, and have some idea if you’re making things better or worse.
Conclusion
None of the above was rocket science, and you don’t need to do some complex FEM analysis to make the average hacker project. But a bit of real engineering can go a long way towards more reliable mechanisms, and that starts with knowing the numbers you’re dealing with. Taking the required measurements can be simple if you know how to build the tools you need, and your life will be easier with some numbers to guide you.
The robotic revolution is currently happening, although for the time being it seems as though most of the robots are still being generally helpful to humanity, whether that help is on an assembly line, help growing food, or help transporting us from place to place. They’ve even showed up in our homes, although it’s not quite the Jetsons-like future yet as they mostly help do cleaning tasks. There are companies that will sell things like robotic vacuum cleaners but [Clay Builds] wanted one of his own so he converted a shop vac instead.
The shop vac sits in a laser-cut plywood frame and rolls on an axle powered by windshield wiper motors. Power is provided from a questionable e-bike battery which drives the motors and control electronics. A beefy inverter is also added to power the four horsepower vacuum cleaner motor. The robot has the ability to sense collisions with walls and other obstacles, and changes its path in a semi-random way in order to provide the most amount of cleaning coverage for whatever floor it happens to be rolling on.
There are a few things keeping this build from replacing anyone’s Roomba, though. Due to the less-than-reputable battery, [Clay Builds] doesn’t want to leave the robot unattended and this turned out to be a good practice when he found another part of the build, a set of power resistors meant to limit current going to the vacuum, starting to smoke and melt some of the project enclosure. We can always think of more dangerous tools to attach a robotic platform to, though.
Conectar
Flippin Rhobot is controlled by an ESP32 that listens into the chat on [DJ Pfeif]’s stream. He’s got a vaguely humanoid form, and he can rotate on the spot and wave his arms in the air courtesy of a few servos. He’s also got a little computer terminal that displays the show’s “Hack the Planet” logo when he turns to face the screen. His body also features some addressable LEDs that flash and dance on command.
