Use Machine Learning To Identify Superheroes And Other Miscellany

December 12, 2016 by Gerrit Coetzee 6 Comments

[Massimiliano Patacchiola] writes this handy guide on using a histogram intersection algorithm to identify different objects. In this case, lego superheroes. All you need to follow along are eyes, Python, a computer, and a bit of machine learning magic.

He gives a good introduction to the idea. You take a histogram of the colors in a properly cropped and filtered photo of the object you want to identify. You then feed that into a neural network and train it to identify the different superheroes by color. When you feed it a new image later, it will compare the new image’s histogram to its model and output confidences as to which set it belongs.

This is a useful thing to know. While a lot of vision algorithms try to make geometric assertions about the things they see, adding color to the mix can certainly help your friendly robot project recognize friend from foe.

TastingFeet: Building Toes And Tongues

December 12, 2016 by Brian Benchoff 4 Comments

Noodle Feet is a robot — an artistically designed robot — that is a character from Sarah Petkus’ webcomic Gravity Road. This webcomic explores a post-human universe inhabited by robots, and dives deep into these robots’ exploration of the trash left behind from a human civilization.

Sarah’s not just drawing these robots. She’s bringing them to life. The character Noodle Feet, so named because his legs are encased in pool noodles, has been made real with an aluminum skeleton, a PCB brain, and infrared detecting eyes. At the 2016 Hackaday SuperConference Sarah gave a talk on the challenges of making this robot real and the specifics of making her robot dig its toes into carpet, slobber all over the floor, and taste with its artificial tongue.

Since last year’s talk on Noodle Feet, Sarah has vastly improved the gripping strength of her noodle’s feet. Over the last two years of construction the mechanism to extend grippy, cat-like toenails has moved from cheap hobby servos to solenoids to a clever cam system. While these toe feet worked, the grip was never quite right, and the world isn’t completely covered in shag carpet. After the break we’ll take a closer look at the improvements that Sarah made to the design and how she came up with the ideas for each new iteration.

Continue reading “TastingFeet: Building Toes And Tongues” →

Train Your Robot To Walk With A Neural Network

December 11, 2016 by Elliot Williams 22 Comments

[Basti] was playing around with Artificial Neural Networks (ANNs), and decided that a lot of the “hello world” type programs just weren’t zingy enough to instill his love for the networks in others. So he juiced it up a little bit by applying a reasonably simple ANN to teach a four-legged robot to walk (in German, translated here).

While we think it’s awesome that postal systems the world over have been machine sorting mail based on similar algorithms for years now, watching a squirming quartet of servos come to forward-moving consensus is more viscerally inspiring. Job well done! Check out the video embedded below.

Continue reading “Train Your Robot To Walk With A Neural Network” →

Animatronic Head Responds With Animated GIFs

December 10, 2016 by Dan Maloney 6 Comments

[Abhishek] describes Peeqo as a “personal desktop robotic assistant” that looks like “the love child of an Amazon Echo and a Disney character.” We’re not sure about that last part — we’re pretty sure [Bender Bending Rodriquez] would fail a paternity suit on this one. Just look at that resemblance.

Whatever Peeqo’s parentage may be, it’s a pretty awesome build, and from the look of [Abhishek]’s design notes, he put a lot of thought into it, and a lot of work too. The build log reveals 3D-printed parts galore, custom-etched PC boards, and a hacked Raspberry Pi to both listen for voice commands and play responses in the form of animated GIFs on Peeqo’s ‘face’. The base has six modified RC servos to run the Gough-Stewart platform that lets Peeqo emote, and the head contains pretty much all the electronics. Beyond the hardware, a ton of programming went into giving Peeqo the ability to communicate through head gestures and GIFs that make sense for the required response.

Whether it’s bopping along to the tunes on your playlist or motivating you to lay off the social media with [Will Ferrell]’s flaming angry eyes, Peeqo looks like a ton of fun to build and use. Conveniently enough, [Abhishek] has shared all his files so you can build one too.

We haven’t seen anything like Peeqo before, but we have seen a lot of Amazon Echo hacks and even a few Stewart platform builds. But did we inadvertently feature a project starring Peeqo’s dad way back in 2009?

[Thanks to Aaron Cofield for the tip]

Speed Run [James Bruton’s] Star Wars Builds

December 10, 2016 by Rich Hawkes 4 Comments

We’ve been following [James Bruton]’s builds here on Hackaday for quite a while and he has built some impressive stuff. We love how he often doesn’t cover everything up, leaving enough room to admire the working bits under the hood. Just in time for the release of the new Star Wars movie, Rogue One, [James] put together an overview of his Star Wars robot builds.

The build summary includes his R6 droid, his GNK walking droid and the third revision of his BB-8 droid. [James Bruton]’s videos have tons of detail in them over many, many parts (for example, his BB-8 R3 playlist is 15 parts and his Ultron build currently has 26 episodes and counting!)

There’s a quick overview of each of the three robot builds in this video, and it includes links to the playlists for each build for those who want more detail. This is just what you need to glimpse all of the clever design that went into these wonderfully crafted droids. And if you haven’t seen it yet, you should check out his series elastic actuators that he’s working on for the Ultron build, they give a robot some relief from rigidity.

Continue reading “Speed Run [James Bruton’s] Star Wars Builds” →

Building A Whegged Robot

December 6, 2016 by Richard Baguley 13 Comments

[Jaidyn Edwards] is building a robot. This isn’t going to be a normal robot, though, he’s building a whegged robot, inspired by Boston Dynamic’s version of the RHex design.

A wheg (TM) is a curved leg that rotates around a ~~foxed~~ fixed (Ed note: Fixed!) point on one end, driven by a motor. Hence the name: part wheel, part leg. By driving each leg separately, you can keep the robot balanced and push it forwards. This is a complex system to build. Unlike normal wheels or drive systems, you need to know exactly where the leg is to use it properly, as the position of the leg depends on the rotation of the motor.

The legs themselves are going to be 3D printed from a combination of rigid and flexible fabrics that should provide both strength and grip. In this first video, [Jaidyn] outlines his design, and explains why he is trying this approach. It’s the first in an ongoing series that should definitely be worth tuning into.

Continue reading “Building A Whegged Robot” →

Otto Bot Is Bob’s Grandson

December 4, 2016 by Elliot Williams 15 Comments

The Otto DIY robot has just taken first place in the coveted role as “best robot to 3D print for your (inner) child”. It’s cute, it dances, it doesn’t cost too much, it’s completely open source, and it’s not impossible to write code for. It’s probably the most refined Bob design that we’ve seen yet. Watch it move in the video below.

Continue reading “Otto Bot Is Bob’s Grandson” →