Research

Pervasive Systems Research Data Sets

Data sets, source codes, and APPs

Pervasive Systems group has a number of data sets, source codes, and APPs, which with a proper acknowledgement, can be used for research purposes.

Multi Sensor-Orientation Movement Data of Goats

  • Download link
  • Citation notice: (a) Jacob W. Kamminga, Duv V. Le, Jan Pieter Meijers, Helena C. Bisby, Nirvana Meratnia, and Paul J.M. Havinga, "Robust sensor-orientation-independent feature selection for animal activity recognition on collar tags", In the Proceedings of ACM Interactive, Mobile, Wearable and Ubiquitous Technologies, Vol. 2, Issue. 1, March 2018, (b) Kamminga, MSc J.W. (Universiteit Twente) (2018): Multi Sensor-Orientation Movement Data of Goats.
  • This is a labeled dataset. Motion data were collected from six sensor nodes that were fixed with different orientations to a collar around the neck of goats. These six sensor nodes simultaneously, with different orientations, recorded various activities performed by the goat. We recorded the activities of five different goats on two farms in the Netherlands. We used a 3-axis accelerometer, high-impact accelerometer, gyroscope, and magnetometer. All sensors were sampled at 100 Hz. For more details please see the readme file in the dataset.

Smoking Activity Dataset

  • Download link (size around 2 GB)
  • Citation notice: (a) Shoaib, Muhammad, Hans Scholten, Paul JM Havinga, and Ozlem Durmaz Incel. "A hierarchical lazy smoking detection algorithm using smartwatch sensors." In 18th IEEE International Conference one-Health Networking, Applications and Services (Healthcom), (b) Shoaib, O. D. Incel, H. Scholten, and P. J. Havinga, “SmokeSense: Online Activity Recognition Framework on Smartwatches,” in 9th EAI International Conference on Mobile Computing, Applications and Services (MOBICASE), FEBRUARY 28–MARCH 2, 2018
  • This dataset contains more than 40 hours of sensor data for smoking and other similar activities. It was collected using a smartwatch at the wrist position and a smartphone in pocket position. The details about the dataset and its collection process is described in the readme file

RSS dataset for indoor localization using smartphones, WiFI access points, and iBeacons

  • RSS dataset for indoor localization
  • Citation notice: (a) Duc V. Le and Paul J. M. Havinga (2017) SoLoc: Self-organizing Indoor Localization for Unstructured and Dynamic Environments. In: The 18th International Conference on Indoor Localization and Indoor Navigation (IPIN 2017) 18-21 November 2017, Sapporo, Japan; (b) This data collector is supported by the Dutch national program COMMIT in the context of the Lost and Found project.
  • This dataset contains WiFi and Bluetooth received signal strength (RSS) measurements of a smartphone at 603 grid positions in entire the DesignLab, Gallery building, University of Twente. Among 603 positions, which can be split for training and testing. This testbed includes 11 WiFi access points and 46 Bluetooth beacons that were deployed on the celling, of which height is about 4-6 meters. This dataset can be used for both fingeringting localization and rang-based localization. Further details can be found in the readme file.

Labeled behavioral data from 4 individual goats and 2 sheep

  • Animal behavioral data
  • Citation notice: (a) Jacob W. Kamminga, Helena C. Bisby, Duc V. Le, Nirvana Meratnia, and Paul J.M. Havinga. Generic online animal activity recognition on collar tags. In Proceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp/ISWC'17). ACM, September 2017, (b) Kamminga, MSc J.W. (University of Twente) (2017): Generic online animal activity recognition on collar tags
  • This dataset comprises labeled behavioral data from 4 individual goats and 2 sheep that exercised 9 different activities. The sensor types included are: accelerometer, gyroscope, magnetometer, temperature, and pressure. The dataset contains 1 day of data for each animal. The sensor units were always placed around the neck of the animals and the orientation was not fixed (the collars were prone to rotation around the neck). All the sensors were sampled with 200Hz. More details can be found within the archive (see Readme.pdf)

Complex Human Activities Dataset (13 activities)

  • Download link (size around 300 MB)
  • Citation notice: (a) Shoaib, Muhammad, Stephan Bosch, Ozlem Durmaz Incel, Hans Scholten, and Paul JM Havinga. "Complex human activity recognition using smartphone and wrist-worn motion sensors", In Sensors 16, no. 4 (2016): 426, (b) Shoaib, M., Bosch, S., Scholten, H., Havinga, P. J., Incel, O. D. (2015), "Towards detection of bad habits by fusing smartphone and smartwatch sensors", In 2015 IEEE International Conference on Pervasive Computing and Communication Workshops (PerCom Workshops)
  • This dataset contains different smartphone sensors data for 13 human activities (walking, jogging, sitting, standing, biking, using stairs, typing, drinking coffee, eating, giving a talk, and smoking). The details about the dataset and its collection process is described in the readme file.

Source code for the simulator used in paper entitled 'Unified routing for data dissemination in smart city networks'

  • Source code simulator
  • Citation notice: (a) Le, Viet Duc and Scholten, J. and Havinga, P.J.M. (2012) Simulation Source Codes of Unified Routing for Data Dissemination in Smart City Networks paper published in (IoT 2012), Version V.21, supported by the SenSafety project in the Dutch Commit program, (b) Le, Viet Duc and Scholten, J. and Havinga, P.J.M. (2012) Unified routing for data dissemination in smart city networks. In: 3rd International Conference on the Internet of Things, IOT 2012, 24-26 Oct 2012, Wuxi, China. pp. 175-182. IEEE Press. ISBN 978-1-4673-1346-9
  • This is the source code that was used to run simulations for results published in the paper, entitled 'Unified routing for data dissemination in smart city networks'. The source code was developed from the open source code The One simulator that can be found at https://www.netlab.tkk.fi/tutkimus/dtn/theone/ . The source code includes not only numerous new java classes but also all setting files for simulations conducted in purpose for the paper.

Physical Activity Recognition Dataset Using Smartphone Sensors

  • Activity recognition dataset
  • Citation notice: (a) Shoaib, M. and Scholten, J. and Havinga, P.J.M. (2013) Towards physical activity recognition using smartphone sensors. In: 10th IEEE International Conference on Ubiquitous Intelligence and Computing, UIC 2013, 18-20 Dec 2013, Vietri sul Mare, Italy. pp. 80-87. IEEE Computer Society. ISBN 978-1-4799-2481-3, (b) This dataset is supported by the Dutch national program COMMIT in the context of the SWELL project.
  • This dataset contains smartphone sensors data for six physical activities. The data was collected using four participants. Moreover, each participant was provided with 4 smartphones on four body positions ( jeans pocket, arm, wrist, belt) so data was collected for each activity on these 4 positions. The activities were walking, running, standing, sitting, and walking upstairs and downstairs. Data was collected for three smartphone sensors (an accelerometer, a gyroscope, a magnetometer) at 50 samples per second. Further details can be found in the readme file in dataset archive.

Sensors activity dataset

Android App for Data Collection

  • Android App for Data Collection
  • Citation notice: (a) Shoaib, M. and Scholten, J. and Havinga, P.J.M. (2013) Towards physical activity recognition using smartphone sensors. In: 10th IEEE International Conference on Ubiquitous Intelligence and Computing, UIC 2013, 18-20 Dec 2013, Vietri sul Mare, Italy. pp. 80-87. IEEE Computer Society. ISBN 978-1-4799-2481-3, (b) This data collector is supported by the Dutch national program COMMIT in the context of the SWELL project.

Sensors Android App for Data Collection