Finding Plastic Spaghetti With Machine Learning

March 29, 2019 by 10 Comments

Among 3D printer owners, “spaghetti” is the common term for the tangled mess of stringy plastic that’s often the result of a failed print. Fear of their print bed turning into a hot plate of PLA spaghetti is enough to keep many users from leaving their machines operating overnight or while they’re out of the house. Accordingly, we’ve seen a number of methods that allow the human operator to watch their print remotely to make sure everything is progressing smoothly.

But unless you plan on keeping your eyes on your phone the entire time you’re out of the house, there’s still a chance some PETG pasta might sneak its way out. Enter the Spaghetti Detective, an open source project that lets machine learning take over when you can’t sit watching the printer all day. Their system plugs into Octoprint to monitor your print in real-time and pause it if it starts looking particularly stringy. The concept is still under development, but judging by the gallery of results submitted by users, the system seems to have a knack for identifying non-edible noodles.

Once the software comes out of beta it looks like the team is going to try to monetize it by providing hosting and monitoring services for a monthly fee, but as it’s an open source project, you’re also able to run the software on your own machine. Though the documentation notes that the lowly Raspberry Pi doesn’t have quite what it takes to handle the image recognition routines, so you’ll need a proper computer if you want to self-host the service. Could be a good use for that old laptop you’ve got kicking around the lab.

As demonstrated in the video after the break, the system’s “spaghetti confidence” is shown with a simple to understand gauge: green is a good-looking print, and red means the detective is getting a sniff of the stringy stuff. If your print dips into the red too much, Octoprint is commanded to pause the print. The user can then look at the last image from the printer and decide to either cancel the print entirely, or resume if the Spaghetti Detective got a little overzealous.

Frankly, it’s a brilliant idea and we’re very interested to see where it goes from here. Assuming you’ve got Octoprint controlling your 3D printer there are some very clever monitoring systems out there currently, but since spaghetti isn’t the only thing a rogue 3D printer can cook up, having an extra line of defense sounds like a good idea to us.

Continue reading “Finding Plastic Spaghetti With Machine Learning” →

How Much Apple Does A Hamburger Get You?

March 29, 2019 by 28 Comments

A while ago, [Skippy] bought a cheap knock-off of the Apple USB mains charger from an AliExpress seller, for the British low, low price of 89p. Normally we’d give you a dollar conversion, but since that’s coincidentally the price of the basic McDonalds hambuger in the UK we’ll go with the hamburger as a unit of conversion. And as any self-respecting hacker would, he subjected it to a teardown and gave it a few tests.

Surprisingly though its pins were a little long it was just within the BS1363 pin spacing specification, probably due to its external dimensions copying the Apple original. The emissions test he performed might surprise readers, as it gave the little device its first pass. Radiated RF emissions were well below the test threshold, a welcome sight for anyone who has had to test a device. Sadly the same could not be said for conducted emissions, and it was happily spraying RF to all and sundry from its connections.

Taking a look inside revealed the usual litany of frightening safety fails. There was no insulation between the mains pins and the circuit board, and a secondary capacitor was even touching one of the pins. Meanwhile another capacitor connecting both sides of the circuit was not of the required Y rating. These and a raft of others make the device illegal for sale in Europe without further tests, but to give some numbers to it all he subjected it to a screen test applying 600 VAC common mode to its pins and checking for leakage current through the device. This it failed, and indeed it did not recover from the test.

So in this case, the price of a hamburger definitely does not get you an Apple, nor even does it get you an equivalent. But of course, you knew that, because we’ve talked about fake Apple chargers and power supplies many times before.

Writing A Very Tiny Chess Program

March 28, 2019 by 29 Comments

When programming for modern platforms, the restraints are different to those of 30 years ago. Back in the dawn of the microcomputer age, storage and RAM were measured in kilobytes. It simply wasn’t possible to store large amounts of graphical data, and even code had to be pared back at times. [reeabgo] found out some of these limitations first hand, when coding a tiny chess program for the Sinclair ZX81.

[reeabgo]’s project goes by the name ChesSkelet, and is truly tiny. Measuring in at just 377 bytes in its smallest version, the entire program takes up less space than this very article describing it. To achieve these feat requires certain sacrifices, of course. The tiniest edition contains no graphics whatsoever, representing the game state with simple characters and featuring no adornments whatsoever. The full-fat version comes in at 477 bytes and adds quite a lot of functionality. There’s a proper checkerboard, along with move legality checks and pawn promotion.

Unfortunately, advanced chess play isn’t quite possible – castling is not implemented, and the AI doesn’t yet handle check situations properly. Despite this, it’s a solid approximation of the real game, all packed into an impressively small space.

We see plenty of chess hacks around these parts – including the robotic variety.

Eavesdropping On Cosmonauts With An SDR

March 28, 2019 by 9 Comments

Usually when we hear about someone making contact with astronauts in orbit, it’s an intentional contact between a ham on the ground and one of the licensed radio amateurs on the ISS. We don’t often see someone lucky enough to snag a conversation between ground controllers and a spacecraft en route to the ISS like this.

For [Tysonpower], this was all about being in the right place at the right time, as well as having the right equipment and the know-how to use it properly. Soyuz MS-12 launched from Baikonur on March 14 with cosmonaut [Aleksey Ovchinin] and NASA astronauts [Nick Hague] and [Kristina Koch] onboard, destined for the ISS after a six-hour flight. The lucky bit came when [Tysonpower] realized that the rendezvous would happen when the ISS was in a good position relative to his home in Cologne, which prompted him to set up his gear for a listening session. His AirSpy Mini SDR was connected to a home-brew quadrifilar helical (QFH) “eggbeater” antenna on his roof. What’s nice about this antenna is that it’s fixed rather than tracking, making it easy to get on the air with quickly. After digging around the aviation bands at about 121 MHz for a bit, [Tysonpower] managed to capture a few seconds of a conversation between [Ovchinin] and Moscow Flight Control Center. The commander reported his position and speed relative to the ISS a few minutes before docking. The conversation starts at about 1:12 in the video below.

We think it’s just cool that you can listen in on the conversations going on upstairs with a total of less than $50 worth of gear. Actually talking to the hams aboard the ISS is another matter, but not a lot more involved really.

Continue reading “Eavesdropping On Cosmonauts With An SDR” →

Reverse Engineering A Modern IP Camera

March 28, 2019 by 51 Comments

Security cameras used to be analog devices feeding back into a room full of tiny screens and commercial grade VCRs. As technology moved forward, IP cameras began to proliferate. Early models simply presented a video stream and configuration page to the local network. Modern models aimed at the home market differ however. More often than not, configuration is through a strange smartphone app, and video is accessed through third-party servers. It’s all a bit oblique, and so [Alex] decided to take a look under the hood. 

The exploration begins externally, with [Alex] capturing data sent to and from the camera with Wireshark. Straight away, red flags are raised. For as yet unknown reasons, the camera attempts to resolve Google, Facebook and Alibaba servers over DNS. Disassembly then follows, revealing that a serial terminal with root access is available. [Alex] uses this to probe around, uncovering the firmware update script and a way to decrypt said updates.

The work thus is a great example of how to approach hacking a given device from first principles. The overall goal is to find a way to gain complete control over the camera, reprogramming it to serve up video as [Alex] wishes, rather than to a distant third party server. It’s not the first time we’ve seen an IP camera hacked, and we doubt it will be the last. If you’ve got one cracked, be sure to let us know.

Ammo Can Holds A 14,000 Lumen LED Flashlight

March 28, 2019 by 10 Comments

For most people, a flashlight is just something you keep in a drawer in the kitchen in case the power goes out. There’s even a good chance your “flashlight” is just an application on your phone at this point. But as we’ve seen many times before from mechanical keyboards to Power Wheels, hardcore niche communities can develop around the most innocuous pieces of hardware; and the lowly flashlight is no different.

Case in point, this 14,000 lumen LED flashlight built by [Bryson Hicks]. Designed around a 100 watt module from Stratus LED, the flashlight uses a number of 3D printed components to make itself at home in a suitably hardcore enclosure: a metal ammo can. With the addition of some modular electronics and a rather slick little control panel, his light is ready to deliver an unreasonable level of brightness anywhere he wishes.

The Stratus LED module includes its own driver, and just needs to be hooked up to a suitably beefy power source to do its thing. [Bryson] went with a 4500 mAh LiPo battery that he says gets him about a one hour runtime at full brightness. For somewhat less intense operation, he’s added a potentiometer which interfaces with the module’s driver board to control the LED output. Considering how fast the light sucks down the juice, adding a small LCD battery charge indicator to the top of the device seems like it was a prudent decision.

To prevent you from cooking anyone’s eyes at close range, the light requires you to first “arm” it by flipping the military style protected switch. Once the switch is in the on position, an illuminated push button is used to actually turn the LED module on and off. You can also snap the toggle switch back into the closed and covered position if you needed to kill the light in a hurry.

This isn’t the first preposterously bright LED flashlight we’ve seen around these parts. There’s something of an arms-race between hackers and makers to develop increasingly bright lights they can carry around, on the off chance they need to illuminate an entire neighborhood.

An Air Quality Monitor That Leverages The Cloud

March 28, 2019 by 9 Comments

Air quality has become an increasing concern in many urban areas, due to congestion and our ever-increasing energy use. While there are many organisations that task themselves with monitoring such data, it’s also something anyone should be able to take on  at home. [Chrisys] is doing just that, with some impressive logging to boot.

The build starts with a Raspberry Pi Zero W, which offers the requisite computing power and Internet connectivity in a compact low-power package. For determining air quality, the Bosch BME680 sensor is used. This offers temperature, pressure, and humidity readings, along with the ability to sense the presence of volatile organic compounds, or VOCs. These can be harmful to human health, so it’s useful to have an idea of the levels in your home.

The hardware is incredibly refined. It’s simple enough for the newbie, but just begs for the more experienced hacker to expand on.

On the software side, data is accessible through the Balena cloud service. Sensor readings are stored in an InfluxDB instance, with Grafana providing the visually attractive graphs and monitoring. It’s all very slick and Web 2.0, and can be accessed from anywhere through a web browser.

The project is a great example of combining a basic DIY Raspberry Pi setup with the right software tools to create a polished and effective end product. Of course, if you’re looking for something more portable, this project might be more your style.