This Robot Can’t Keep Its Eyes Off The Money

November 7, 2021 by Zoe Skyforest 6 Comments

Some say there’s no treasure quite as valuable as the almighty dollar. [Norbert Zare] likes alt-rock soundtracks on Youtube videos and robots obsessed with money, so set about building the latter.

The project is fundamentally a simple one. A Raspberry Pi 3B+ is outfitted with a Pi Camera, and set up to control twin servo motors attached to a simple pan/tilt assembly. The Pi runs OpenCV set up in a face-tracking mode. This allows the robot to readily track money in its field of view, as the vast majority of money out there has someone’s face on it. OpenCV is used to detect where the money is in the field of view, and guide the Pi’s camera towards the cash.

It’s a neat repurposing OpenCV’s face detection algorithm, and that’s much faster than training your own money-tracking system. However, it seems like the robot would also track regular human faces, too. Perhaps it could be optimised to do a color check, such that only greyscale or green faces were followed by the robot.

Does the project do anything useful or important? Arguably no, but if a robot can be this obsessed with money, perhaps we all can learn something. Alternatively, it might just have served as a useful project for [Norbert] to learn about programming and mechatronics projects. Either way, we dig it. Code is on Github for the curious.

Using OpenCV in this way has become common over the years. If you want to detect cats, however, maybe consider giving Tensorflow a try. Video after the break.

Continue reading “This Robot Can’t Keep Its Eyes Off The Money” →

Homebrewed Voice Assistant Keeps An Eye On Air Quality

November 5, 2021 by Zoe Skyforest 3 Comments

Voice assistants are now available from a wide variety of companies, however, [7402] didn’t like the idea of these devices sending data off to the cloud for potentially-nefarious purposes. Thus, the goal became to build a home voice assistant that worked entirely offline, and that’s precisely what [7402] achieved.

The system had limited goals compared to commercial competitors. [7402] was more than happy to deal with a limited vocabulary of understanding as a trade off for privacy. It’s all built around a Raspberry Pi Zero, which runs the Julius speech recognition library. Ultrasonic sensors are used to only activate the device when a person leans in and directly addresses the system.

Capabilities include reporting on the weather, switching light on and off, and advising users of air quality readings from the local authorities. Feedback to the user is via text-to-speech as well as flashing LEDs. The latter are used to create a quirky, retro “thinking” animation to indicate the system is processing, and has indeed heard a spoken command.

It’s a neat build, and one that covers most of the good things that commercial cloud devices are capable of anyway. As a bonus, no smartphone apps are required, nor will private companies impact the system’s functionality as it relies on no external servers to operate.

We’ve seen similar builds before too, such as this GlaDOS-themed voice assistant. Video after the break.

Continue reading “Homebrewed Voice Assistant Keeps An Eye On Air Quality” →

Battle Robot Uses Carbon Fiber To Save Weight

November 5, 2021 by Zoe Skyforest 9 Comments

ZAP! The saw is capable of delivering high-voltage discharges to damage its foes.

Combat robots come in all shapes and sizes, with regulating authorities often using weight limits to create a level playing field for competitors. [Hans Jørgen Grimstad] is building a robot to compete in a 4 kg class, and made some interesting design decisions to that end.

4 kg is not a lot of weight to play with. When considering the motors needed to propel the robot and the batteries needed to run everything, there’s then precious little weight left for weapons systems and armor plating.

Thus, in an effort to make the most of the weight limitations, [Hans] decided to use carbon fiber for the robot’s outer shell. The method used is a simple wet layup in a mold. We’d be supremely interested to see how this armor holds up in competition, versus more typical choices like aluminium and steel.

Other interesting features include a belt-driven saw, which [Hans] tests with his hands mere inches away and the robot’s motors powered up. Don’t do this if you value your fingers. This is paired with a high-voltage discharge taser module. When the saw gets close to another robot, it may cause sparks to jump to the enemy, damaging its electronics in the process. It’s something we haven’t seen too often, as such measures are actually banned in some contests.

Diehard enthusiasts in the battle robot community will likely have fierce opinions on many points of the design; have it out in the comments. It’s certainly not the first carbon-fiber bot, but it’s nice to see the fancy material being thrown in the ring.

We’ve seen other designers innovate, too, such as this remarkably successful walking robot build. Video after the break.

Continue reading “Battle Robot Uses Carbon Fiber To Save Weight” →

Hacking The Mekamon Robot To Add New Capabilities

November 3, 2021 by Zoe Skyforest 6 Comments

The Mekamon from Reach Robotics is a neat thing, a robot controlled by a phone app that walks on four legs. [Wes Freeman] decided to hack the platform, giving it a sensor package and enabling some basic autonomous behaviours in the process.

[Wes] started out by using a packet sniffer to figure out the command system for controlling the Mekamon robot over Bluetooth. Then, he set about fitting a Raspberry Pi 3 on the ‘bot, along with a Pi Camera on a gimballed camera head.

Running OpenCV on the Raspberry Pi gives the Mekamon robot the ability to follow a colored ball placed in its field of vision. Later work involved upgrading the hardware to a Pi Compute Module 3, with its dual camera inputs allowing for the use of a stereo imaging setup.

All the parts simply ziptie on top of the original robot, with no permanent changes needed. It’s a neat way of hacking, by expanding the original capabilities without actually having to tamper within.

We’ve seen plenty of autonomous builds over the years, from farming robots to those designed to explore the urban environment. Video after the break.

Continue reading “Hacking The Mekamon Robot To Add New Capabilities” →

How To Get Into Cars: Endurance Racing Builds

November 2, 2021 by Zoe Skyforest 21 Comments

Many an automotive enthusiast finds themselves at a track day eventually. Typically, this involves competing against the clock to better one’s laptimes in short sessions throughout the day. Such events are fun, but it often creates a perishing thirst for a greater level of competition.

Regularity and endurance events are often the next step up for a lot of people. These events involve long runs at race pace that stress a car to (or beyond!) the breaking point. Careful preparation is required if one is to see out the race to the chequered flag. Let’s break down what you’ll need to consider.

Continue reading “How To Get Into Cars: Endurance Racing Builds” →

Hedgehog Gesture Sensor Built With Cheap Time-of-Flight Modules

November 2, 2021 by Zoe Skyforest 6 Comments

Time-of-flight sensors used to be expensive obscurities, capable of measuring the travel time of photons themselves and often used for tracking purposes. However, the technology is cheaper now, such that [jean.perardel] has used TOF sensors to build a useful and affordable gesture-tracking system.

The system relies on four VL53L1X time of flight sensors, which have a 16×16 scanning array and communicate over the I2C bus. Controlling the show is an Arduino MKR1010, though the project should be achievable with a range of other microcontrollers, too.

The device is built into a cute hedgehog-like form factor, with an LCD screen acting as the face. It displays facial expressions which show how the system is interpreting and responding to gestures. It gives the project lots of personality, which makes using the system more fun. Gestures from the system can be used to send keystrokes over USB, control relays or servos, or even fire IR signals to control TVs and other hardware.

It actually seems like a useful gesture control interface, one that could become a useful part of a workstation setup. We’ve seen gesture controls put to other uses too, like controlling robot arms. Video after the break.

Continue reading “Hedgehog Gesture Sensor Built With Cheap Time-of-Flight Modules” →

Supersonic Baseball Hitting A Gallon Of Mayo Is Great Flow Visualization

November 1, 2021 by Zoe Skyforest 22 Comments

Those of us who enjoy seeing mechanical carnage have been blessed by the rise of video sharing services and high speed cameras. Oftentimes, these slow motion videos are heavy on destruction and light on science. However, this video from [Smarter Every Day] is worth watching, purely for the fluid mechanics at play when a supersonic baseball hits a 1-gallon jar of mayo.

The experiment uses the baseball cannon that [Destin] of [Smarter Every Day] built last year. Ostensibly, the broader aim of the video is to characterize the baseball cannon’s performance. Shots are fired with varying pressures applied to the air tank and vacuum levels applied to the barrel, and the data charted.

However, the real glory starts 18:25 into the video, where a baseball is fired into the gigantic jar of mayo. The jar is vaporized in an instant from the sheer power of the collision, with the mayo becoming a potent-smelling aerosol in a flash.

Amazingly, the slow-motion camera reveals all manner of interesting phenomena. There’s a flash of flame as the ball hits the jar, suggesting compression ignition happened at impact with the jar’s label. A shadow from the shockwave ahead of the ball can be seen in the video, and particles in the cloud of mayo can be seen changing direction as the trailing shock catches up.

The slow-motion footage deserves to be shown in flow-visualization classes, not only because it’s awesome, but because it’s a great demonstration of supersonic flow phenomena. Video after the break.

Continue reading “Supersonic Baseball Hitting A Gallon Of Mayo Is Great Flow Visualization” →