Data for: Office activity recognition using accelerometers and gyroscopes located on the forearms

Published: 29 June 2019| Version 1 | DOI: 10.17632/2ws2g3pn6w.1
Jose Alejandro Avellaneda Gonzalez, Ivan David Alfonso Diaz, Monica Tatiana Gutierrez Ballen


This dataset contains information of gyroscopes and accelerometer located on forearms meanwhile 5 activivities are performed (Use the computer, Have a phone conversation, Take notes-handwriting, Drink a Beverage and Look at the smartphone). These activities are typical in an office. Sensors collect data during 3 minutes approximately when the person seats down and develops each specific activity. There is a csv file per user-activity-trial which contains records of the 3 minutes activity. There are 230 csv files in total which have information of 7 users, the users 1-4 performed 5 trials each one and users 5-7 performed 1 trial. - The activities are numbered in the csv files and their names as follows: 1. Use the computer 2. Drink a Beverage 3. Have a phone conversation 4. Take notes-handwriting 5. Look at the smartphone - The columns name are related in the headerLA (letf forearm) and headerRA (rigth forearm) files.


Steps to reproduce

Hold two smartphones in each forearm, ask the user to perform the activities described below and start to record the data once the user begins to perform the activity. 1. Use the computer: login his email and write an email of approx 200 words or 3 minutes. 2. Drink a Beverage: take a sip, count 3 seconds and take another sip. This routine should last 3 minutes. 3. Have a phone conversation: Have a phone conversation for a maximum of 3 minutes 4. Take Notes: write 20 times a sentence of 6 words. 5. Use a smartphone: Look at the smartphone, unlock it and check social networks during 3 minutes.


Activity Recognition, Machine Learning, Sensor, Office Ergonomics