Windows IoT uses cognitive API to identify callers and then opens a door with a Raspberry Pi

Masato Sudo, a software engineer in the Microsoft Windows IoT team has developed a system which recognises a person approaching a door and then unlocks (or doesn’t) accordingly.

A standard USB camera captures the image, the Pi takes the image in and, using the Microsoft Cognitive Services API, identifies the person.

“It’s a prototype that identifies the visitor.” If the door recognises the visitor, it will greet them by name, and the door will be unlocked and opened.

If they aren’t identified, the door remains firmly locked. The opening of the door is controlled by a GPIO pin that sends a signal to a relay. The code is C# and was written using Visual Studio.

You can read more here and see a full tutorial of how he did it, including the code, here.

Light framework for reading and sending sensor data written in C++/C# for the Raspberry Pi

Achilleus from Samnium has contacted me about an application he’s written for the Raspberry Pi 3.

It is a light framework for reading data from multiple sensors and sending them over the network. It’s called IoTWork.Reader
and was developed in C++ and C#; it can be executed using Mono.

The code is available on GitHub. The current release supports the SenseHAT and Achilleus has made pre-packaged binaries which are also located on GitHub.

The application is customizable: you can write your own sensor handlers and configure them via the XML file.

More documentation is available on GitHub.

French Raspberry Pi website containing tutorials, news and downloads

Pierre-Lin Bonnemaison has been in touch. For almost three years, he has been working on two websites for the Raspberry Pi. The first, Raspbian France is in the French language and has various tutorials and news as well as links to downloads and an active forum. The second, How To Raspberry Pi, is a translation of the first site into English. He’s done a really good job on putting the site together and it’s great to see resources written in a language other than English. So, if you speak French, head over to Raspbian France or How To Raspberry Pi if you speak English.

Mycroft – an open source artificial intelligence voice processor for the Raspberry Pi

Artificial intelligence software Mycroft was originally a Kickstarter. Designed as an open source alternative to Siri, Google Now, Cortana and Echo, it was originally developed for a Raspberry Pi 2 and is hosted on GitHub. Now, the Mycroft team has released an SD card image for the system for Raspberry Pi 2 & 3 so you can ‘talk to your Pi’ without needing to do anything apart from write a card. You can read more and download the image here and add a speaker and USB microphone to make your voice-processing artificial intelligence Pi come to life.

Pimoroni solderless headers for the Raspberry Pi reviewed

Jenny List over at Hackaday recently posted about the Pimoroni solderless headers. Met with a certain amount of resistance and skepticism from the Hackaday crowd as to the suitability of both the method of hammering in the pins and their longevity, she decided to conduct an experiment. Following the hammering of the pins to secure them, she removed and replaced a Gert VGA board from/to the pins over a hundred times. She found that the pins were absolutely fine and solid after the experiment, proving, at least in a slightly unscientific way, that the method is sound. Read more over at Hackaday and see a speeded-up video of the experiment below. The kit to fit the solderless headers, and the headers themselves are available from Pimoroni.

A project to build a Raspberry Pi Zero-driven mobile phone

=

There have been several Raspberry Pi-powered mobile phone builds over the past few years (Tyler Spadgenske and David Hunt) but none have used the Pi Zero… until now. (Cue dramatic music).

Arsenijs from Latvia has created the ZeroPhone, and he reckons it has only cost around $50 to build! It can be used for making calls, sending SMS messages and has various text-based mini-apps pre-loaded. It runs Raspbian and pyLCI for the UI elements. The bill-of-materials is as follows:

  • Pi Zero – easy to get despite people on Internet still telling otherwise (I got 5 by now). Let’s say it’s 5$ and 5$ for shipping.
  • SIM800 modules – available in large quantities from eBay, 5$.
  • ESP8266-12E used for WiFi (as per RPi-WiFi project) – 2$ from eBay
  • Two-layer PCBs (two 4x10cm boards, one 4x6cm board) – can be ordered from your favourite fab or whatever people on Internet give good reviews about.
  • ATMega328P – can be desoldered from an Arduino Pro Mini that’s typically 2$ on eBay, and it comes with a bootloader already! If you don’t have a heat gun, you can just buy a bare chip and burn a bootloader to it using a programmer.
  • LCD screen – 5$ on eBay. Besides, a project like this can use other screen breakouts with similar characteristics (such as the ubiquitous Nokia displays) – it’s just a matter of writing a driver, which is easier than it sounds, and the front board accepts different display pinouts.
  • Battery – you can use cell phone batteries, pouch and 18650 cells.
  • TP4056 battery charger – modules are available from eBay, 1$.
  • Buttons for keypad – 2$ from eBay (there are 3 types used, but you can omit one).
  • 2.54 headers (both dual-row and single-row) – dirt cheap. Won’t be more than 0.50$ to have enough to assemble the phone.

It’s currently at working prototype stage and is, basically, a Pi Zero sandwiched between circuit boards. You can read more over on Hackaday.io.