Just picked up one of these neat ePaper displays from AliExpress for $9.90 for the esp8266. It’s 2.13" in diagonal, has a resolution of 250x122 and is made by Lolin.

Lolin produce a whole bunch of quality esp8266 items - including the ubiquitous Wemos D1 Mini and various I2C sensors. It’s not a shield in that you can sit it on top of an ESP - you need a couple of cables to use the lcd with the esp, you’ll want to get the TFT I2C Connector Shield ($0.70) and TFT e-Paper Cable ($0.80).

A minor annoyance, but one that plagues me: when you put a wash on, forget about it finishing, only to discover the damp washing hours later festering in the washing machine. There’s only one thing to it - a massively overengineered smart home solution!

Sonoff POW

This is a smart device that can be had from Chinese online retailers for ~$9 - bargain. In it there’s a wifi microcontroller, the venerable ESP8266, a mains relay and a current/voltage sensor to measure power usage. It’s the power monitoring functionality this hack uses.

Problem: a recipe calls for 250ml double cream (48% fat) and 100ml whole milk (3.5% fat). I only have double cream (48% fat) and semi-skimmed milk (1.5% fat).

How much do I adjust the double cream by to get the same quantity liquid and fat content?

It’s a simple simultaneous equation, so solving it with Sympy is overkill, but developers do overkill best, right?

So:

>>> linsolve([Eq(x + y, 350), Eq(x*.48 + y*.015, 250*.48 + 100*.035)], (x, y))

ie. volume must be 350ml, fat must equal the original recipe.

This is part 2 of 2 about using Elasticsearch and Kibana to track patterns in baby activity. Part 1 here covers the hardware and setup for tracking baby kicks.

Machine learning

Having collected about 6 weeks of baby kicking data, it’s time to test the new toy in the Elasticsearch stack: Machine learning.

Installing this was a straightforward case of following the instructions, and from the ‘Machine Learning’ new menu item in Kibana, I chose ‘Create new job’, and ‘Create a single metric job’:

This is part 1 of 2 about using Elasticsearch and Kibana to track patterns in baby activity.

Part 2 is here.

Kicking things off

According to countthekicks.org “Counting baby kicks is important because changes in your baby’s movement pattern may indicate potential problems with your pregnancy”. Counting and patterns sounds like a technological problem for Elasticsearch and Kibana!

I struggled getting the Yale Conexis smart lock working with the Aeotec Z-Stick Gen5. Turns out the issue is it must be a secure pairing - the default method in the instructions (unplug Z-Stick, press button near to the device, initiate pairing), does not perform a secure pairing, and so even though the device will appear in software, you won’t get any state for it or be able to perform an actions.

The new Raspberry Pi Zero W is the perfect pairing to go with the Xiaomi Mi Flora. The Flora is a bluetooth stick you push into the soil next to a plant and can be polled over Bluetooth LE for the sensor data - soil moisture, temperature, light and soil conductivity.

Part list

• Raspberry Pi Zero W. I used Pimoroni (£9.60 + £2.50 postage) as they’re great and ship fast. You shouldn’t need the Adapters kit if you’ve the necessary adapters already, or you might get away with setting up the Pi headless without them anyway.

Bluetooth doesn’t work out the box for Archlinux but thankfully a helpful user has created packages to support the Pi 3 which also work on the Zero W (see forum post). There’s a couple of caveats to getting them working I’ll step through.

First install bluez and bluez-utils older than 5.44 - you must get them from an archive. The reason being hcitool is required and it’s been dropped from bluez 5.44.

This project was to build a simple garage security system to alert of intruder access to the garage by notification over Wifi.

Parts list

Wemos D1 mini & mini protoboard

The esp8266 is the workhorse of any project - it monitors the sensors and sends notifications over Wifi. I used the Arduino system with Platformio for development.

For the small amount of circuitry required one of these tiny protoboards:

The Amazon Dash is a Wifi button originally intended for the ultra- lazy for reordering consumables from Amazon when you’re running low. It was going to £1 recently in Black Friday promotions, so I thought I’d have to snap one up. Normally they go for a not bank breaking £5, which you can recoup if you actually want to order one of the items once on them.

When pressed it connects through Wifi to send a request to Amazon, then turns off - looks fit for repurposing for home automation!