I just recently got some smart Wi-Fi Plugs with energy monitoring to see what is causing my bill to go 20-30% higher.
Now, I'm thinking about using ML.NET to optimize my power consumption based on their energy usage and my habits.
The project is intended to be fun and improving things on a small scale. 🏞
Hopefully, this will end up being a project that anyone can pick up and not only save some money in the long run but also help the environment by collectively be more efficient with the energy. 🌏
I don't have a team.
If anyone interested in home automation or trying to figure out how to make schedules based on machine learning, you're welcome to join. 🤗
Brief Description
My goal is to strike a good balance between always-on and completely off with help of automation and how the device handles shutdowns. ⚡
For example, we don't want to hard power down the PC while it's working. When we are at home, we are more likely to use the PC, so it should be in a low-power state and go fully off (when in hibernation, can we trigger hibernation somehow?) when we go to sleep or outside. Similarly, with consoles like Xbox, we want to be able to fully turn off but also take into consideration that the user might not want to lose game progress, giving them some "always-on" grace period before the console goes off and all progress not saved is lost. (Yes, Xbox One Series X solve this problem but many people don't have it 😊)
The goals would be:
Detect devices states.
Hibernations (PC: ~3Wh)
Hybrid sleep (PC: ~7Wh, Xbox ~11Wh)
Idle (PC: 110Wh, Xbox: ~40Wh)
Normal usage (PC: ~150Wh, Xbox: ~50Wh)
Gaming (PC: ~300Wh, Xbox: ~70Wh)
Determine when it's safe to turn off the device
For PC when in hibernation (<5Wh for more than a minute)
For Xbox when I'm ok to lose progress? (sleep with or without game running looks the same)
For streaming devices when TV is off
What else we can turn off to save energy?
Track habits and determine a good time to turn off the devices
Machine Learning that adapts over time based on usage per device
Hackathon Idea
I just recently got some smart Wi-Fi Plugs with energy monitoring to see what is causing my bill to go 20-30% higher. Now, I'm thinking about using ML.NET to optimize my power consumption based on their energy usage and my habits. The project is intended to be fun and improving things on a small scale. 🏞
Hopefully, this will end up being a project that anyone can pick up and not only save some money in the long run but also help the environment by collectively be more efficient with the energy. 🌏
Your name
Jernej Kavka (JK) - https://github.com/jernejk
Team name
I don't have a team. If anyone interested in home automation or trying to figure out how to make schedules based on machine learning, you're welcome to join. 🤗
Brief Description
My goal is to strike a good balance between always-on and completely off with help of automation and how the device handles shutdowns. ⚡
For example, we don't want to hard power down the PC while it's working. When we are at home, we are more likely to use the PC, so it should be in a low-power state and go fully off (when in hibernation, can we trigger hibernation somehow?) when we go to sleep or outside. Similarly, with consoles like Xbox, we want to be able to fully turn off but also take into consideration that the user might not want to lose game progress, giving them some "always-on" grace period before the console goes off and all progress not saved is lost. (Yes, Xbox One Series X solve this problem but many people don't have it 😊)
The goals would be:
I'm using TP-Link Kasa HS110 for gathering energy consumption. https://www.tp-link.com/au/home-networking/smart-plug/hs110/
Other
Are you looking for team members?
Would you like to have a mentor assigned to your team?