A few years ago, I started experimenting with home automation. Not because I wanted to be the smartest person on the block, but because I was tired of paying too much for electricity every month — being away during the day and running everything at peak hours in the evening.
Now my home runs mostly on autopilot. The lights, the EV charger, the heat pump — everything communicates. And the bill is significantly lower. Here’s how I did it.
Why standard timers aren’t enough
A timer is better than nothing, but it doesn’t adapt to reality. What if you come home early? What if the sun is shining and you don’t need lights at all? What if electricity prices turn negative in the middle of the day?
That’s why I use an AI-driven system that makes continuous decisions based on three factors:
- Real-time electricity price (via the EPEX day-ahead market)
- Weather (cloudy → more indoor lighting needed)
- My presence (GPS from my phone)
The system also learns. After a few weeks, it knows when I usually come home, whether I sleep in on weekends, and when I’m on vacation.
Smart lights: the low-hanging fruit
Let’s start with lighting — the cheapest and easiest part of home automation.
I use Philips Hue in the living room and Ikea Tradfri everywhere else. Both are open enough to control via Home Assistant or n8n. The price difference is significant — Ikea costs about a third of Hue — but the quality is perfectly fine.
What does it save?
- Lights only turn on when someone is in the room
- In the morning, they dim in gradually instead of flashing on
- Outdoor lights are based on sunset, not a fixed time
- In summer, they stay off more often because it stays light longer
Savings: about €50-80 per year in direct electricity. That doesn’t sound groundbreaking, but it’s the easiest part. The real gains come from combining it with a home battery and dynamic energy tariffs.
The home battery: buy power when it’s cheap
This is where it gets interesting. I have a 10 kWh home battery — a lithium-ion unit about the size of a small washing machine. It charges when electricity is cheap (or negative) and discharges when prices are high.
Even without solar panels, you can profit from this. The principle is simple:
- At 2:00 AM, electricity is cheapest → charge the battery
- At 6:00 PM, electricity is most expensive → discharge the battery
With a dynamic contract (I use Zonneplan and ANWB), I pay the hourly rate, not a fixed tariff. The difference between off-peak and peak can be 20 to 30 cents per kWh.
Math: 10 kWh × €0.25 difference × 300 days = €750 per year.
On top of that, I also use the battery to store my own solar power. On sunny days, the battery charges during the day, and I use that power in the evening instead of expensive grid electricity.
The AI that runs it all
Without automation, I’d have to manually switch between charging and discharging. I don’t — I simply forget. So everything runs through an AI-driven system I built in n8n.
The workflow works like this:
- Every hour, a webhook fetches the current electricity price
- The system checks the weather forecast (sun → extra yield)
- It checks my calendar (am I home or not?)
- Based on this data, it controls the battery: charge, discharge, or wait
The logic isn’t complicated, but the combination of factors makes it powerful. A simple example:
It’s 2:00 PM, the electricity price is €0.05/kWh, the sun is shining, and I’m not home. The AI decides: leave the battery idle and feed solar power directly to the grid. Why? Because the battery is already full and it’s better to sell now at a reasonable price.
Two hours later, 4:00 PM. The price has dropped to €0.02. The AI decides: charge the battery a little from the grid, so it can discharge at maximum during tonight’s peak.
The system re-evaluates these decisions every few minutes. A human can’t keep up with that manually, but a computer can.
What does it cost?
The investment isn’t negligible, but it pays for itself:
| Component | Cost | Payback period |
|---|---|---|
| Smart lights (entire house) | €400 | 4-5 years |
| 10 kWh home battery (installed) | €4,500 | 5-6 years |
| AI system (n8n, server, sensors) | €300 | 1-2 years |
| Total | €5,200 | 4-6 years |
After that, it’s pure profit. And the battery lasts at least 10 years (under warranty), so you get another 4-6 years of extra benefit.
Want to do the same?
You don’t need a technical background to get started. The easiest entry point: buy a few smart lights (Ikea Tradfri, no hub needed with Zigbee2MQTT) and sign up for a cheap dynamic energy contract. Track the data in a spreadsheet — you’ll see the difference within a month.
Want to go further? Install Home Assistant on a Raspberry Pi or an old laptop. That’s the beating heart of a smart home. Then build the AI logic in n8n to control everything.
Have questions or want advice on which system suits your situation? Get in touch — I’m happy to help.