Dataset Open Access

Long-Term Tracing of Indoor Solar Harvesting

Sigrist, Lukas; Gomez, Andres; Thiele, Lothar

Dataset Information

This dataset presents long-term term indoor solar harvesting traces and jointly monitored with the ambient conditions. The data is recorded at 6 indoor positions with diverse characteristics at our institute at ETH Zurich in Zurich, Switzerland.

The data is collected with a measurement platform [3] consisting of a solar panel (AM-5412) connected to a bq25505 energy harvesting chip that stores the harvested energy in a virtual battery circuit. Two TSL45315 light sensors placed on opposite sides of the solar panel monitor the illuminance level and a BME280 sensor logs ambient conditions like temperature, humidity and air pressure.

The dataset contains the measurement of the energy flow at the input and the output of the bq25505 harvesting circuit, as well as the illuminance, temperature, humidity and air pressure measurements of the ambient sensors. The following timestamped data columns are available in the raw measurement format, as well as preprocessed and filtered HDF5 datasets:

  • V_in - Converter input/solar panel output voltage, in volt
  • I_in - Converter input/solar panel output current, in ampere
  • V_bat - Battery voltage (emulated through circuit), in volt
  • I_bat - Net Battery current, in/out flowing current, in ampere
  • Ev_left - Illuminance left of solar panel, in lux
  • Ev_right - Illuminance left of solar panel, in lux
  • P_amb - Ambient air pressure, in pascal
  • RH_amb - Ambient relative humidity, unit-less between 0 and 1
  • T_amb - Ambient temperature, in centigrade Celsius

The following publication presents and overview of the dataset and more details on the deployment used for data collection. A copy of the abstract is included in this dataset, see the file abstract.pdf.

L. Sigrist, A. Gomez, and L. Thiele. "Dataset: Tracing Indoor Solar Harvesting." In Proceedings of the 2nd Workshop on Data Acquisition To Analysis (DATA '19), 2019.

Folder Structure and Files

  • processed/ - This folder holds the imported, merged and filtered datasets of the power and sensor measurements. The datasets are stored in HDF5 format and split by measurement position posXX and and power and ambient sensor measurements. The files belonging to this folder are contained in archives named yyyy_mm_processed.tar, where yyyy and mm represent the year and month the data was published. A separate file lists the exact content of each archive (see below).
  • raw/ - This folder holds the raw measurement files recorded with the RocketLogger [1, 2] and using the measurement platform available at [3]. The files belonging to this folder are contained in archives named yyyy_mm_raw.tar, where yyyy and mmrepresent the year and month the data was published. A separate file lists the exact content of each archive (see below).
  • LICENSE - License information for the dataset.
  • README.md - The README file containing this information.
  • abstract.pdf - A copy of the above mentioned abstract submitted to the DATA '19 Workshop, introducing this dataset and the deployment used to collect it.
  • raw_import.ipynb [open in nbviewer] - Jupyter Python notebook to import, merge, and filter the raw dataset from the raw/ folder. This is the exact code used to generate the processed dataset and store it in the HDF5 format in the processed/folder.
  • raw_preview.ipynb [open in nbviewer] - This Jupyter Python notebook imports the raw dataset directly and plots a preview of the full power trace for all measurement positions.
  • processing_python.ipynb [open in nbviewer] - Jupyter Python notebook demonstrating the import and use of the processed dataset in Python. Calculates column-wise statistics, includes more detailed power plots and the simple energy predictor performance comparison included in the abstract.
  • processing_r.ipynb [open in nbviewer] - Jupyter R notebook demonstrating the import and use of the processed dataset in R. Calculates column-wise statistics and extracts and plots the energy harvesting conversion efficiency included in the abstract. Furthermore, the harvested power is analyzed as a function of the ambient light level.

Dataset File Lists

Processed Dataset Files

The list of the processed datasets included in the yyyy_mm_processed.tar archive is provided in yyyy_mm_processed.files.md. The markdown formatted table lists the name of all files, their size in bytes, as well as the SHA-256 sums.

Raw Dataset Files

A list of the raw measurement files included in the yyyy_mm_raw.tar archive(s) is provided in yyyy_mm_raw.files.md. The markdown formatted table lists the name of all files, their size in bytes, as well as the SHA-256 sums.

Dataset Revisions

v1.0 (2019-08-03)

Initial release.
Includes the data collected from 2017-07-27 to 2019-08-01. The dataset archive files related to this revision are 2019_08_raw.tar and 2019_08_processed.tar.
For position pos06, the measurements from 2018-01-06 00:00:00 to 2018-01-10 00:00:00 are filtered (data inconsistency in file indoor1_p27.rld).

v1.1 (2019-09-09)

Revision of the processed dataset v1.0 and addition of the final dataset abstract.
Updated processing scripts reduce the timestamp drift in the processed dataset, the archive 2019_08_processed.tar has been replaced.
For position pos06, the measurements from 2018-01-06 16:00:00 to 2018-01-10 00:00:00 are filtered (indoor1_p27.rld data inconsistency).

Dataset Authors, Copyright and License

References

[1] L. Sigrist, A. Gomez, R. Lim, S. Lippuner, M. Leubin, and L. Thiele. Measurement and validation of energy harvesting IoT devices. In Design, Automation & Test in Europe Conference & Exhibition (DATE), 2017.

[2] ETH Zurich, Computer Engineering Group. RocketLogger Project Website, https://rocketlogger.ethz.ch/.

[3] L. Sigrist. Solar Harvesting and Ambient Tracing Platform, 2019. https://gitlab.ethz.ch/tec/public/employees/sigristl/harvesting_tracing

Appears in the Proceedings of the 2nd Workshop on Data Acquisition To Analysis (DATA '19)
Files (158.8 GB)
Name Size
2019_08_processed.files.md
md5:90564ce317367d0557ba61684fb5602f
1.8 kB Download
2019_08_processed.tar
md5:2fa48880fc47b15d513bdcc8b838e9ea
32.5 GB Download
2019_08_raw.files.md
md5:fded5d19604139033a4bfa480dd7b5bf
286.5 kB Download
2019_08_raw.tar
md5:7d99c4674d31decc9b34c5f4163620cb
126.3 GB Download
abstract.pdf
md5:59ad173cbe56082622e7da9466f4c213
919.3 kB Download
LICENSE
md5:0d7ac17eb2b045281c5f26cab21c4b2d
338 Bytes Download
processing_python.ipynb
md5:1127a733a17a2a5428a75f1e887a8662
142.9 kB Download
processing_r.ipynb
md5:90276ad17de59da36d9c02629b6bea4b
425.9 kB Download
raw_import.ipynb
md5:9c736f7bc157318f04fa3eb654052de4
20.9 kB Download
raw_preview.ipynb
md5:3373f7a034fd430f840c879a12372b7b
306.6 kB Download
README.md
md5:30540309497c092eb501f9e3f9de432c
7.7 kB Download
174
268
views
downloads
All versions This version
Views 174127
Downloads 268170
Data volume 6.1 TB4.0 TB
Unique views 136103
Unique downloads 9066

Share

Cite as