Creating my HAL9000 openHAB system

Hi everyone,

Just joined this community and posted my first question. Thought it would be fun to introduce me and my system too. Hope you share that idea :wink: I’m a bit of a gadgeteer so I’ve gone a bit wild. The heart is the usual Raspberry Pi but I combined it with some electronics to be able to signals over internet cables (the usual PWM, Digital, Analogue). Some of them are combined with internet (of course losing speed to 100 mbit because you steal 4 wires). To prevent chaos I reused a broken switch which has all of the required connectors and makes up for a nice casing. Since there was room to spare I also included a harddrive for a NAS. Last but nog least using the USB it’s connected to the power meter and Zigbee devices. As a trusty component it should have a trusty name, so I called it the HAL9000 :smiley:

And placed in the electricity cabinet it looks like this

It’s use mostly by voice command but to have some extra fun I also included a dashboard with graphs and buttons. Maybe inevitable because I work as a BI specialist…

So now I working my way through openHAB to expand the system. It’s not always as easy as I hoped but I found some great examples on this forum. And created my first widget and used open API’s to get the weather and the buss timetable close to my home on the dashboard. A little less open was getting the data from my solar system (envoy), but on internet I found a guy who hacked to local API from the device. Worked like a charm. And of course, all that programming needs some soldering to compensate. So using ESP’s I created a MP3 doorbell and multi-light switch that the HAL can control.

Some of my solutions looks a bit over-complicated, so maybe more questions will come. But I’m also the type, if it works it works. Well enough for now, I had some fun writing this down. No idea if these openHAB stories are actually read…

9 Likes

Seems like a risky name for a home automation system.

“Open the front door HAL.”

“I’m sorry, I can’t do that Michel.”

Welcome to the community!

7 Likes

They are not only read but appreciated! Bring it on!
I’m interested in your BI work there, mind sharing some details??

Hi @Pedro_Liberal ! Well, a bit (maybe a lot :slight_smile: ) off-topic. I’m working in the government creating information about projects, think hours and money spent and results accomplished (milestones/ user stories completed or incidents solved). But to see trends and relations it’s a lot of visuals, graphs and the meters are great for KPI’s or SLA’s. So that reminded me while making the openHAB dashboard. But of course it’s all enterprise tools like Powercenter (ETL) and Business Objects. And Java, Knime, R-Studio for the things need to be derived from patterns. And Kibana for the real-time information. No openHAB in the office :wink:

Maybe I’ll use this knowledge more in the future. For now I was really happy things showed on the dashboard in de tiles I made for them. Layout and getting what I wanted needed some searching on f7-framework and Apache echarts. The standard components don’t seem to work well on small screens. A whole new world besides openHAB, fortunately I was not the only one. Also lot’s of great topics about that on this forum. Glad I know CSS though. But you have to be real accurate in the yaml to get in working. Anyway … enough for your curiosity? (be assured no :cat: were killed for this)

1 Like

Yep! Thanks for indulging my curiosity! I’m working on my home graphs and it’s nice to see what is being done elsewhere. One thing I started doing was adding sensors everywhere. Light sensors near windows, and balconies, temp. / humidity sensor per room, now I’m looking to add pm2.5 sensors in frequently used rooms and CO in my office and kitchen…
For what? Not 100% sure yet! But by having their data in my database I can refer back to it in months or years time for research.
I love that!!

My kind of guy. I’ve got the stuff ready to make ESP cheap I2C sensors connected by WiFi to the openHAB server. The big plan is to use the Nibe binding to get the heatpump data and combine with weather and sunlight data. A windy day with clouds must take more power to heat than a sunny day without wind. With enough historical data I must be able to make a model and show actual against predicted data. More fun to come :tada:

1 Like