TensorFlow Robot Recognizes Objects

Children can do lots of things that robots and computers have trouble with. Climbing stairs, for example, is a tough thing for a robot. Recognizing objects is another area where humans are generally much better than robots. Kids can recognize blocks, shapes, colors, and extrapolate combinations and transformations.

Google’s open-source TensorFlow software can help. It is a machine learning system used in Google’s own speech recognition, search, and other products. It is also used in quite a few non-Google projects. [Lukas Biewald] recently built a robot around some stock pieces (including a Raspberry Pi) and enlisted TensorFlow to allow the robot to recognize objects. You can see a video of the device, below.

Continue reading “TensorFlow Robot Recognizes Objects”

Slide Viewer Upgrade is a Bright Idea

[cunningfellow] has been putting LEDs in everything lately. That’s understandable. Most recently, he used them to drastically upgrade his father’s super-cool mid-century slide viewer.slide-viewer-comparison

The slide viewer used to use a flashlight bulb, but it didn’t light evenly at all. Not only that, it produced a dim, orange-ish light. [cunningfellow] happened to have an old Nokia N93 lying around and decided to cannibalize that strange, beautiful, swiveling flip phone for its backlight circuitry. Unfortunately, the 4 LEDs aren’t going to run on a pair of C cells like the flashlight bulb did. [cunningfellow] needed some kind of boost converter.

He found one in the form of a Nokia E73 LCD driver board created by [Andy Brown]. The LEDs are way brighter than that old incandescent bulb, and they draw about 10mA less to boot. We think [cunningfellow]’s father will be happy with the result.

If you have an old slide viewer and no slides, try using it as a project case. If this post makes you miss your View-Master (also understandable), you can always turn your phone into a stereopticon.

Hackaday Links: October 9, 2016

Atari is back! That’s what some dude says. There are no real details in that post, other than ‘Atari is Back!’

The ESP32 is coming, and it’s going to be awesome. Espressif has just released an Arduino core for the ESP32 WiFi chip. The digitalRead, digitalWrite, SPI, Serial, Wire, and WiFi “should” work. If you’re looking for ESP32 hardware, they’re infrequently available and frequently out of stock. Thankfully, stock levels won’t be the Raspberry Pi Zero all over again until someone figures out how to run an NES emulator on the ESP32.

Tiny, cheap ARM boards would make for great home servers if they had SATA or multiple network interfaces. Here’s a Kickstarter for a board with both. It’s based on an ARM A53 with multiple Ethernets, mini PCIe, enough RAM, and SATA. It’s a board for niche use cases, but those uses could be really cool.

You’re not cool or ‘with it’ until you have a PCB ruler. That’s what all the hip kids are doing. For wizards and dark mages out there, a simple PCB ruler isn’t enough. These rare beasts demand RF rulers. There’s some weird stuff on these rulers, like Archemedian spiral antennas and spark gaps. Black magic stuff, here.

Some dude with a camera in the woods did something. Primitive Technology, the best example of experimental archaeology you’ve ever seen, built a spear thrower. You can throw a ball faster with a lacrosse stick than you can with just your hands, and this is the idea behind this device, commonly referred to as an atlatl. You can hunt with an atlatl in some states, but I have yet to see a video of anyone taking down a deer with one of these.

Think we’re done spamming the Hackaday Superconference yet? YOU’RE WRONG. The Hackaday Superconference is the greatest hardware conference of all time until we do this whole thing again next year. Get your tickets, look at the incredible list of speakers, book your flights, and be in Pasadena November 5-6.

Arcade Cabinet Build Takes Quarters, Dispenses Fun

Building an arcade cabinet seems to be a rite of passage for many hackers and woodworkers. Not that there is anything wrong with that: as this series of posts from [Alessandro] at boxedcnc shows, there is an art to doing it well.

His final build is impressive, with quality buttons, a genuine-looking banner, and even a coin slot so he can charge people to play. His build log covers both the carpentry and electronic aspects of the build, from cutting the panels to his own code for running the coin acceptor that takes your quarter (or, as he is in Italy, Euro coins) and triggers the game to play.

To extract money from his family, he used the Sparkfun COM-1719 coin acceptor, which can be programmed to send different pulses for different coins, connected to an Arduino which is also connected to the joystick and buttons. The Arduino emulates a USB keyboard and is connected to an old PC running MAME with the Attract Mode front end. It’s a quality build, down to the Bubble Bobble banner, and the coin slot means that it might even make some money back eventually.

Cheap 360 Degree Camera

How much would you pay for a 360 degree camera? How about $15 if you already have a Raspberry Pi and a Pi camera hanging around? If you don’t, you’ll have to add that minimal cost into the build. [Gigafide] noticed how a spherical mirror, made to see around corners, showed an all-around view if you took a picture of it from below. He snagged a panoramic lens made for an iPhone and stripped it for its optics. Some custom software and a little work resulted in a usable 360 degree camera.

SimpleCV (a light version of OpenCV) provides the algorithms to unwrap the frames and you can take video with the setup (see the video below). Mounting the optics took some 3D printing and the Pi operates as a hot spot to send the video out.

Continue reading “Cheap 360 Degree Camera”

Simulate Your Robot Before You Build It

[Nurgak] shows how one can use some of the great robotic tools out there to simulate a robot before you even build it. To drive this point home he builds the tutorial off of the easily 3D printable and buildable Robopoly platform.

The robot runs on Robot Operating System at its core. ROS is interesting because of its decentralized and input/output agnostic messaging system. For example, if you leave everything alone but swap out the motor output from actual motors to a simulator, you can see how the robot would respond to any arbitrary input.

[Nurgak] uses another piece of software called V-REP to demonstrate this. V-REP is a simulation suite for robotics and has a few ROS nodes built in. So in order to make a simulated line-following robot, [Nurgak] tells V-REP to send a simulated camera image to the decision making node of the robot in ROS. It then sends the movement messages back to V-REP which drives the pretend robot around.

He runs through a few more examples, proving that it’s entirely possible to become if not a roboticist, at least a really good AI programmer without ever dropping the big money on parts to build a robot.