Collect, evaluate and predict Activities of Daily Living from within Home Assistant.
Activities of Daily living (ADLs) such as eating, working, sleeping and Smart Home device readings are recorded by inhabitants. Predicting the activities of residents based on the device event-stream allows for a range of applications, including automation, action recommendation and abnormal activity detection in the context of assisted living for elderly inhabitants. Activity Assistant is a platform that streamlines the data collection process. The plugin extracts device readings from Home Assistant and tracks residents using an additional Android or the Home Assistant companion app. Moreover, Activity Assistant (will) support the deployment of trained models and running ADL predictions in real-time. A (future) Home Assistant integration provides users with a novel activity-based abstraction for automating their homes.
The installation of this add-on is pretty straightforward and not different to installing any other Hass.io add-on.
- Add the repository to the addon-store by pasting
https://github.com/tcsvn/hassio-activity-assistantinto the manage add-on repositories dialog - Ensure that Home Assistant is using the recorder integration for Activity Assistant to successfully connect to the database and query devices
- Search for the "Activity Assistant" addon-on in the Hass.io add-on store and install the latest release (NOT edge or development).
- Start the "Activity Assistant" add-on.
- Check the logs of the add-on to see if everything went well.
- Ready to go!
For a guide on how to use please refer to the Documentation (tbd)
- Create and run experiments for recording ADLs and device events
- Activity annotation using an android app
- Activity annotation using Home Assistant input_selects and input_booleans
- Device or activity to room assignment
- Interactive dashboard for the running experiment and previously recorded datasets
- Prior activity distributions with a "typical week" calendar widget
- Dask nodes for distributed computation
- Upload and deployment of trained models via the web interface
- Home Assistant integration + custom entity card displaying per person activity predictions
- Fork it (https://github.com/tcsvn/activity-assistant/fork)
- Create your feature branch (
git checkout -b feature/fooBar) - Commit your changes (
git commit -am 'Add some fooBar') - Push to the branch (
git push origin feature/fooBar) - Create a new Pull Request
- Activity Assistant - Logger - The android companion app for labeling activities
- pyadlml - A python library offering data exploration methods for ADL datatasets.
- Home Assistant - Homeautomation platform
If you are using Activity Assistant for puplications please consider citing the package.
@software{activity-assistant,
author = {Christian Meier},
title = {Activty Assistant},
url = {https://github.com/tcsvn/activity-assistant},
version = {0.0.4.1-alpha},
date = {16-02-2023}
}
MIT © tcsvn