A citizen science project by RWTH Aachen University

Donate your smart home’s electricity data to research

Your home often already knows when the heat pump kicks in, when the car charges, and what the solar roof produces. Exactly this knowledge helps researchers design tomorrow’s power grid. With a few clicks in Home Assistant you can contribute your data securely.

SAFEr Grid roadmap: from today's synchronous power grid to asynchronously coupled grid sections
Get started To the data portal

A power grid that works like the Internet

The power grid is facing its biggest rebuild in a century. Until now, a few large power plants have kept everything in lockstep: their heavy turbines act like a flywheel that absorbs every fluctuation. With solar, wind, and batteries, that flywheel dwindles, while electricity flows become more erratic and more local.

The SAFEr Grid research project therefore pursues an idea that has proven itself in the Internet: instead of one rigid whole, many grid sections that balance themselves and pass energy on to each other, much like routers pass on data packets. Four research groups at RWTH Aachen University and Aalborg University are working on the foundations, funded by a Synergy Grant of the European Research Council.

More about the vision at safergrid.energy

Diagram of the SAFEr Grid vision: asynchronously coupled grid sections exchanging energy

Why we need real household data

Whether such a grid stays stable is also decided in households: when thousands of EV chargers start in the evening, when solar roofs feed in at noon while nobody is home, or when people adapt their consumption habits to dynamic electricity prices. Before building a radically new grid, simulations help. But those simulations are only as good as the data they are fed with.

Synthetic load profiles capture the everyday life of individual households only coarsely, and they largely date from the era of large turbine power plants. The dishwasher after the garden party, the heating element in January, the vacation week in August: much of that is often already sitting in your smart home’s records. Donated, it makes the simulations realistic on which the future grid is tested.

A day of electricity decomposed into appliances

The data in question may look like as follows. A household’s grid connection point, next to two individual appliances. The sum curve contains the fridge’s regular consumption as well as a washing program in the evening. Recognizing and modeling individual appliances within the sum is one of the research questions behind the Datenspende.

Scroll to zoom, drag to pan, double-click for the full day. Sample data for illustration.

What you donate

If you run Home Assistant, it already logs how much electricity your devices use and produce. You only share what exists anyway, and you choose which devices take part.

What goes in

  • Power and energy time series (for some devices also voltage and current) of the devices and household readings you approve, such as the heat pump’s daily curve or the solar system’s feed-in.
  • The device type and the device details from your smart home (name, vendor, model), so researchers know what a curve represents: washing machine, EV charger, fridge.
  • Optional, if you choose to provide them: postcode, country, number of people in the household, and construction year of the building if known. This helps us put the data into context.

What stays out

  • Your name, address, precise location, and anything that would identify you directly.
  • Cameras, microphones, presence and motion sensors. They are never read in the first place.
  • No new hardware needed either.

How to participate

1

Install the integration

Add the Datenspende integration to your Home Assistant with a few clicks.

2

Review and consent

See exactly which devices and measurements would be shared, and choose what you are comfortable with.

3

Let it run

From then on, the data flows automatically in the background. There is nothing more to do.

4

Follow your contribution

The data portal shows your donated data, and you can label your devices there, which helps the research further.

Using openHAB, OpenEMS, evcc, openWB, or Shelly? These platforms are on our roadmap.

Read the guide

The donation in numbers

–
Datapoints donated
–
Devices contributing

Donated datapoints over time

Last updated: …

To the data portal

What becomes of your data

An open dataset

Your donation becomes part of an anonymized dataset that we will publish openly and then keep up to date, so energy researchers worldwide can work with it.

Realistic grid simulations

Real household data lets future power grids be simulated as they would actually be used. That is the yardstick for whether the SAFEr Grid concepts hold.

Your own consumption in view

In the data portal you can explore your own consumption and production over time.

Open research

Methods, results, and the dataset itself are shared with the research community. Publicly funded means publicly usable.

Frequently asked questions

Is my data anonymous? What about my privacy?
We collect only electricity-related time series, device details (type, name, vendor, model) and, if you provide them voluntarily, coarse details such as a postcode. We never collect your name, address, or precise location. Your donation is stored pseudonymously: access is through a username and password you choose yourself. That lets us link the donation to your account, so you can view it and have it deleted, but not to you as a person, unless you choose to leave an e-mail address for contact. We are aware that electricity data can reveal habits, such as when someone is home. Therefore we keep the time series separate from data that could point to your identity or where you live. Publication is anonymized only. You decide device by device what gets shared, and the processing follows the GDPR. Details are in our privacy policy. The Datenspende is additionally covered by an approved ethics application of the cross-faculty Ethics Committee of RWTH Aachen University.
Can I stop participating?
At any time, with no reason needed. Simply disable or remove the integration and no further data is transmitted. On request we delete your already donated data from our systems and from future releases of the dataset. To do so, explicitly delete your data in the portal or write us an e-mail. Already published, anonymized releases, however, we can no longer remove.
Can someone access my smart home through the Datenspende?
No. The connection is strictly one-way: the integration sends measurements to our servers and receives no commands. Remotely controlling your devices or intervening in your smart home is technically not possible through the Datenspende. You can find the code on GitHub
Why can I provide an e-mail address?
If you like, you can leave an e-mail address so we can reach you about your donation, for example with questions about your devices or important project changes. The address initially stays linked to your donation, is never published, and is not used for anything else. The donation works just the same without one.
What does it cost, and what do I need?
Participation is free. You need a running Home Assistant installation, an Internet connection, and at least one device that measures power or energy, such as a smart plug, an energy meter, a solar inverter, or an EV charger. If your Internet connection has no flat rate (mobile or similar), costs for the transferred data may arise.
Which platforms are supported?
Home Assistant, currently. openHAB, OpenEMS, evcc, openWB, and Shelly are on the roadmap. If you have suggestions for further platforms or can provide the corresponding interfaces, get in touch at datenspende@comsys.rwth-aachen.de
Who can access the data?
The anonymized dataset will be openly accessible. Our partners in the SAFEr Grid consortium (two further RWTH institutes and Aalborg University) receive pseudonymized data only under data use agreements, without contact data and without a right of onward transfer. For further research questions that need finer-grained data, we plan vetted access for verified researchers under clear usage terms. In the open dataset the time series stay at full resolution while the descriptive details are anonymized: if too few households share the same combination of features, we only pass them on in generalized form, for example as a shortened postcode or a coarser household size bracket (details).
What is the data used for?
For simulating and evaluating future power grids that, for example, must stay stable without the mechanical inertia of large power plants and communicate actively. Real household data makes those simulations more meaningful.
Do I need to be a customer of a specific utility?
No. The Datenspende is independent of energy providers, meter operators, and manufacturers. If your smart home measures electricity and is among the supported platforms, you can take part.
How much bandwidth and storage does it use?
Very little. The integration transmits small, compressed packets of measurements in the background, comparable to a messaging app. You will not notice it in everyday use. If you are not on a flat rate (mobile or similar), however, we advise against participating.
Who is behind this?
The Chair of Communication and Distributed Systems (COMSYS, Prof. Klaus Wehrle) at RWTH Aachen University. The Datenspende is part of SAFEr Grid, a joint project with two further RWTH institutes and Aalborg University.
How is this funded?
Through a Synergy Grant of the European Research Council (ERC), the EU’s funding body for frontier research. The results, including the dataset, will be shared publicly.

From our research

A selection of work related to the Datenspende and SAFEr Grid.

Semi-Intrusive Load Monitoring via Smart Plug Data Fusion for Robust NILM

J. Breyer, L. Pompe, K. Gützlaff, N. Bollwerk, M. H. Alizai, K. Wehrle

SIAM International Conference on Data Mining (SDM), 2026

Resilient low-inertia power systems through asynchronous energy balancing

S. Schwarz, S. Sahoo, M. Stoffers, P. Pollok, L. Pompe, A. M. Fibich, A. Kaharević, K. Wehrle, A. Monti, F. Piller, F. Blaabjerg

Nature Reviews Electrical Engineering, 2026 , doi:10.1038/s44287-025-00256-5

Easing the Conscience with OPC UA: An Internet-Wide Study on Insecure Deployments

M. Dahlmanns, J. Lohmöller, I. B. Fink, J. Pennekamp, K. Wehrle, M. Henze

ACM Internet Measurement Conference (IMC), 2020 , doi:10.1145/3419394.3423666

A citizen science project by RWTH Aachen University

Get started