Software Architecture Laboratory



Related Areas: Ambient Assisted Living, Ambient Intelligence, Applied Research

In the last years, research on techniques able to classify activities of daily living and to detect falls is very active. Many application domains will benefit from these techniques, first of all health. For example, these techniques could help detecting Parkinson’s worsening, calorie expenditure as a result of some sports activities, or falls.
Human activities and falls are recognized analyzing data sampled from sensors such as accelerometer and gyroscope, which are nowadays the basic sensors of each smartphone.
To be effective, detection and classification techniques should be tested and trained with datasets of samples. Unfortunately, publicly available datasets are few, thus, from one hand, it is difficult to make comparative evaluations among the techniques, and from the other hand, researchers are required to waste time in developing ad-hoc applications to sample and label data.
The aim of our work is to provide the scientific community with a system able to ease both the acquisition and the labeling task. The application is Android based and is able to adapt both to the running environment and to the desired protocols the subject has to execute. Due to its simplicity and accuracy, our suite will increase the number of publicly available datasets.

To obtain a free copy of the suite, please send an e-mail to the following address:

If you will use tour app, please cite this paper (downloadable from here):

AUTHOR = {Ginelli, Davide and Micucci, Daniela and Mobilio, Marco and Napoletano, Paolo},
TITLE = {UniMiB AAL: An Android Sensor Data Acquisition and Labeling Suite},
JOURNAL = {Applied Sciences},
VOLUME = {8},
YEAR = {2018},
NUMBER = {8},
URL = {},
ISSN = {2076-3417}

People: Davide Ginelli Paolo Napoletano Marco Mobilio Daniela Micucci

--> © 2024 Software Architecture Laboratory