Energy Analytics Dashboard for Tapo / Kasa | Details

Key functions and features ⚡

Fully automated deployment using Docker Compose, real-time statistics visible immediately
Grafana dashboard pre-configured with the following real-time and long-term metrics:
- Power Usage (W):
  - Histogram plot
  - Numerical readouts for last, highest, lowest, range, mean and total usage
- Voltage (V):
  - Histogram plot
  - Numerical readouts for last, highest, lowest, range, and mean voltage
- Running Cost ($):
  - Average cost per hour/day, estimated total cost
Filter data by MAC and time window, and optionally device name (as configured in Tapo app)
Real-time and historic statistics with instant and projected costs based on trends of the selected time window
InfluxDB v2 back-end time-series optimised database providing hugely efficient storage and indexing; low storage requirements and SQL-like querying
Automatic device discovery, querying and logging using python-kasa and influxdb-client-python
Set and forget with new devices appearing automatically and years of data retention, even on limited storage
Low resource consumption means you can run it on super constrained hardware such as Raspberry Pi 3

Power profiling examples and use cases 📈

Observe and identify specific device power consumption trends over time
Understand duty cycles and determine efficient running modes
Identify inefficient appliances and quantify standby power usage
Track ongoing running costs of specific devices over short and long term
Extrapolate hourly and daily running cost estimates from selected time window
Monitor seasonal or time-of-day usage patterns

Requirements 🔧

Linux host (x86 or ARM): Including Raspberry Pi. Windows is not supported; python-kasa requires host networking mode for broadcast discovery, which isn't available under WSL.
Docker Engine: If not already installed, follow the platform-relevant install instructions here: https://docs.docker.com/engine/install/
Docker user permissions: Ensure your user is added to the docker group (log out/in for changes to take effect):
```
sudo groupadd docker
sudo usermod -aG docker $USER
```
Ssupported TP-Link smart plug(s): One or more Tapo P110 and/or Kasa HS110. Other energy-monitoring models may work but are untested.
Third-Party Compatibility enabled in Tapo app: In the Tapo app, open the account (“Me”) page, navigate to “Third-Party Services”, and ensure that “Third-Party Compatibility” is enabled

Secrets and authentication 🔑

There are six files under /secrets that can be updated; these contain placeholder values only. These are only required when you run docker compose and can be deleted once the container has been fully deployed. Recreate them should you need to re-deploy the container.

InfluxDB's token format is undocumented but appears to accept a base64 string (except using _ and - instead of / and +) up to 88 characters long. Keep the stock token for testing or update if running long-term. If you update the token, it must also be updated in /provisioning/datasources/influxdb.yaml.

Quick start 🚀

Check the Requirements section above and ensure that these are met before proceeding.

Clone the repository to your device:

git clone https://github.com/cjastone/tplink-powerdash.git

Update contents of /secrets if you're not deploying for short-term testing only.

Deploy the container:

docker compose up -d

Deployment time varies; should be under 30 seconds with relatively modern hardware and a fast connection to pull the images, or up to several minutes on heavily constrained hardware.

Once docker compose is finished, wait a few moments for services to start. Optionally watch for issues via docker logs -f telemetry-logger. Initial errors while services are initialising are expected; it will keep retrying until the database exists.

If all went to plan, you should be able to browse to Grafana at http://your.docker.host:3000 and log in using the credentials previously configured under /secrets. Data should be visible immediately.

Troubleshooting 🤔

Running docker logs -f telemetry-logger is your best start - this will give information on database connectivity (especially on first run) as well as device discovery and readings.

Database connectivity issues are visible in Grafana under Connections > Datasources > influxdb-v2 > Test.