ESP8266 And Alexa Team Up To Tend Bar

After a hard day of soldering and posting memes online, sometimes you just want to yell at the blinking hockey puck in the corner and have it pour you out a perfectly measured shot of your favorite libation. It might not be the multi-purpose robot servant we were all hoping to have by the 21st century, but [Jake Lee] figures it’s about as close as we’re likely to get for under fifty bucks or so (Jake’s security certificate seems to have expired a few days ago so your browser may warn you, here’s an archived version).

From the hardware to the software, his Alexa-enabled drink pouring machine is an exercise in minimalism. Not that there’s anything wrong with that, of course. The easiest solutions are sometimes the best ones, and we think the choices [Jake] made here strike a perfect balance between keeping things simple and getting the job done. It’s by no means the most complete or capable robotic bartender we’ve ever seen, but it’s perhaps the one most likely to be duplicated by others looking to get in on the voice-controlled drinking game.

So how does it work? For one, [Jake] didn’t go through the trouble of creating a “proper” Alexa skill, that’s quite a bit of work just to pour a shot of rum. Instead, he took the easy way out and used the FauxMo library on his ESP8266 to emulate a few WeMo smart switches. Alexa (and pretty much every other home automation product) has native support for turning these on and off, so with the proper code you can leverage it as an easy way to toggle the chip’s digital pins.

Using the Alexa’s “Routines” capability, these simple toggles can be chained together and associated with specific phrases to create more complex actions. For example, you could chain the dispensing alcohol, lowering the room lighting, and playing music all to a single voice command. Something like “I give up”, perhaps.

When Alexa tells the drink dispenser to turn on, the ESP8266 fires a relay which starts up a small 12 V air pump. This is connected to the bottle of rum though a glass tube that [Jake] bent with a blow torch, and starts to pressurize it. With the air at the top of the bottle pushing down on it, a second glass tube gives the liquid a way to escape. This method of dispensing liquid is not only easy to implement, but saves you from having to drink something that’s passed through some crusty eBay pump.

If you prefer the “right” way of getting your device talking to Amazon’s popular home surveillance system, our very own [Al Williams] can get you headed in the right direction. On the other hand, if the flowing alcohol is the part of this project that caught your attention, well we’ve got more than a few projects that cover that topic as well.

Picovoice Puts Smarts Offline in 512K of Memory

We live in the future. You can ask your personal assistant to turn on the lights, plan your commute, or set your thermostat. If they ever give Alexa sudo, she might be able to make a sandwich. However, you almost always see these devices sending data to some remote server in the sky to do the analysis and processing. There are some advantages to that, but it isn’t great for privacy as several recent news stories have pointed out. It also doesn’t work well when the network or those remote servers crash — another recent news story. But what’s the alternative? If Picovoice has its way, you’ll just do all the speech recognition offline.

Have a look at the video below. There’s an ARM board not too different from several we have lying around in the Hackaday bunker. It is listening for a wake-up phrase and processing audio commands. All in about 512K of memory. The libraries are apparently quite portable and the Linux and Raspberry Pi versions are already open source. The company says they will make other platforms available in upcoming releases and claim to support ARM Cortex-M, Cortex-A, Android, Mac, Windows, and WebAssembly.

Continue reading “Picovoice Puts Smarts Offline in 512K of Memory”

Voice Controlled Camera for Journalist in Need

Before going into the journalism program at Centennial College in Toronto, [Carolyn Pioro] was a trapeze performer. Unfortunately a mishap in 2005 ended her career as an aerialist when she severed her spinal cord,  leaving her paralyzed from the shoulders down. There’s plenty of options in the realm of speech-to-text technology which enables her to write on the computer, but when she tried to find a commercial offering which would let her point and shoot a DSLR camera with her voice, she came up empty.

[Taras Slawnych] heard about [Carolyn’s] need for special camera equipment and figured he had the experience to do something about it. With an Arduino and a couple of servos to drive the pan-tilt mechanism, he came up with a small device which Carolyn can now use to control a Canon camera mounted to an arm on her wheelchair. There’s still some room for improvement (notably, the focus can’t be controlled via voice currently), but even in this early form the gadget has caught the attention of Canon’s Canadian division.

With a lavalier microphone on the operator’s shirt, simple voice commands like “right” and “left” are picked up and interpreted by the Arduino inside the device’s 3D printed case. The Arduino then moves the appropriate servo motor a set number of degrees. This doesn’t allow for particularly fine-tuned positioning, but when combined with movements of the wheelchair itself, gives the user an acceptable level of control. [Taras] says the whole setup is powered off of the electric wheelchair’s 24 VDC batteries, with a step-down converter to get it to a safe voltage for the Arduino and servos.

As we’ve seen over the years, assistive technology is one of those areas where hackers seem to have a knack for making serious contribution’s to the lives of others (and occasionally even themselves). The highly personalized nature of many physical disabilities, with specific issues and needs often unique to the individual, can make it difficult to develop devices like this commercially. But as long as hackers are willing to donate their time and knowledge to creating bespoke assistive hardware, there’s still hope.

Continue reading “Voice Controlled Camera for Journalist in Need”

Vintage Rotary Phone Turned Virtual Assistant

Like many of us, [Zoltan Toth-Czifra] has completely embraced 21st century living. His home is awash in smart gadgets and dodads, from color changing light bulbs to Internet-connected cameras. But he’s also got a soft spot for the look and feel of vintage hardware, like the rotary phone he keeps kicking around to remind him of the old days. He recently decided to bridge these two worlds by turning the rotary phone into a modern voice controlled assistant.

The first piece of the puzzle was getting the old school phone connected to something a bit more modern, namely a Raspberry Pi. He didn’t want to hack the vintage phone apart, so he picked up a Grandstream HT801, an adapter that’s used to convert analog telephones to VoIP. [Zoltan] says this model specifically fit the bill as it had a function that allows you to configure a number to dial as soon the phone is lifted off the hook. This allows the user to just pick up the phone and start talking without having to dial anything manually. If you’re looking to pull off a similar setup, you should check to make sure the adapter has this function before pulling the trigger.

With the rotary phone now talking a more modern protocol, [Zoltan] just needed to get the Raspberry Pi side sorted out. He installed a SIP server so it could communicate with the HT801 adapter, and then got to work putting together his virtual assistant. Rather than plug into an existing system, he rolled his own by combining open source packages for controlling his various smart devices with the aptly named SpeechRecognition library for Python.

Right now he’s only programmed a few commands that his system can respond to for controlling his lights and music, but mentions that the system is modular enough that he can add new functions easily. He’s put the source for his virtual assistant framework up on GitHub, which he notes was written in less than 200 lines of original code by virtue of utilizing existing libraries for a lot of the heavy lifting. Open source is a beautiful thing.

In the past we’ve seen rotary phones go mobile thanks to GSM upgrades and dragged kicking and screaming onto the modern phone network with a built-in Raspberry Pi. But we think there’s something especially appealing about the approach [Zoltan] took which preserves the phone’s original hardware.

Continue reading “Vintage Rotary Phone Turned Virtual Assistant”

Voice Controlled Stereo Balance With ESP8266

A stereo setup assumes that the listener is physically located between the speakers, that’s how it can deliver sound equally from both sides. It’s also why the receiver has a “Balance” adjustment, so the listener can virtually move the center point of the audio by changing the relative volume of the speakers. You should set your speaker balance so that your normal sitting location is centered, but of course you might not always be in that same position every time you listen to music or watch something.

[Vije Miller] writes in with his unique solution to the problem of the roving listener. He’s come up with a system that can adjust the volume of his speakers without having to touch the receiver’s setup, in fact, he doesn’t have to touch anything. By leveraging configurable voice control software running on his computer, his little ESP8266-based devices do all the work.

Each speaker has its own device which consists of a NodeMCU ESP8266 and X9C104 digital potentiometer inside of a 3D printed case. The audio terminal block on the gadget allows him to connect it inline between the speaker and the receiver, giving [Vije] the ability to adjust the volume through software. The source code, which he’s posted on the Hackaday.io project page, uses a very simple REST-style API to change speaker volume based on HTTP requests which hit the ESP8266’s IP address.

The second part of the project is a computer running VoiceAttack, which lets [Vije] assign different actions based on what the software hears. When he says the appropriate command, the software goes through and fires off HTTP requests to the nodes in the system. Everything is currently setup for two speakers, but it shouldn’t be too difficult to expand to more speakers (or even rooms) with some adjustment to the software.

It’s not the first voice controlled speaker we’ve ever seen, but it does solve a very specific problem in a unique way. We’d be interested in seeing the next logical step, which would see this technology integrated into the speaker itself.

Continue reading “Voice Controlled Stereo Balance With ESP8266”

Alexa And Particle Modernize Coffee Machine By One Iota

When [Steve Parker]’s girlfriend got a tea kettle that takes voice commands, he suddenly saw his fancy bean-to-cup coffee machine as a technological dinosaur. It may make good coffee, but getting the DeLonghi going is inconvenient, because it runs a self-cleaning cycle each time it’s turned on or off.

Thus began [Steve]’s adventure in trying to turn the thing on with Alexa via Particle Photon. Because of the way the machine is designed, simply adding a relay wouldn’t do—the machine would just turn off and back on, only to start the self-clean again. Once inside, he found it’s controlled by a PIC18LF2520. Further research indicated that it is powered by an off-line switcher that combines a power MOSFET with a power supply controller. [Steve] figured out that the buttons are read via square wave and interpreted by a multiplexer.

The project went into the weeds a bit when [Steve] tried to read the signals with a knock-off Saleae. As soon as he plugged it in, the control board fried because the DeLonghi evidently has no reference to Earth ground. While waiting for a replacement board to arrive, he tried replacing the mux and shift register chips, which actually fixed the board. Then it was more or less a matter of using the DeLonghi’s status LEDs to determine the machine’s state, and then to interface with the Photon and Alexa. Cycle past the break for a ristretto-sized demonstration.

[Steve] didn’t do all this to actually make coffee, just turn the machine on with a voice command. The Photon is totally capable of making coffee, though, as we saw with this closed-loop espresso machine.

Continue reading “Alexa And Particle Modernize Coffee Machine By One Iota”

AlterEgo Listens to Your Internal Voice

Recent news reports have claimed that an MIT headset can read your mind, but as it turns out that’s a little bit of fake news. There is a headset — called AlterEgo — but it doesn’t actually read your mind. Rather, it measures subtle cues of you silently vocalizing words. We aren’t sure exactly how that works, but the FAQ claims it is similar to how you experience reading as a child.

If you read much science fiction, you probably recognize this as subvocalization, which has been under study by the Army and NASA. However, from what we know, the positioning of sensor electrodes is crucial and can vary not only by speaker, but also change for the same speaker. Perhaps the MIT device has found a way around that problem. You can see a video of the system, below.

Continue reading “AlterEgo Listens to Your Internal Voice”