TY - GEN
T1 - Classifying Teachers’ Self-reported Productivity, Stress and Indoor Environmental Quality Using Environmental Sensors
AU - Kallio, Johanna
AU - Vildjiounaite, Elena
AU - Kyllönen, Vesa
AU - Ronkainen, Jussi
AU - Koivusaari, Jani
AU - Muuraiskangas, Salla
AU - Räsänen, Pauli
AU - Similä, Heidi
AU - Vehmas, Kaisa
N1 - Funding Information:
Acknowledgements. The authors are grateful to the pilot participants for their active and significant role in the pilot. Special thanks to Mr. Jari Rehu, Mr. Severi Olsbo and our former colleague Mr. Dan Bendas. The study was implemented in collaboration with ESTABLISH and SCOTT projects. ESTABLISH (ITEA3 15008) was supported by Business Finland and VTT Technical Research Centre of Finland, and the pilot was designed in collaboration with UniqAir, CGI, and InspectorSec. SCOTT (Secure COnnected Trustable Things) has received funding from the Electronic Component Systems for European Leadership Joint Undertaking under grant agreement No 737422. This Joint Undertaking receives support from the European Union’s Horizon 2020 research and innovation programme, and from Austria, Spain, Finland, Ireland, Sweden, Germany, Poland, Portugal, Netherlands, Belgium and Norway.
Publisher Copyright:
© Springer Nature Switzerland AG 2019.
PY - 2019/11
Y1 - 2019/11
N2 - Considering that urban people spend majority of their time indoors, buildings should support health and productivity. However, quite commonly unsatisfactory Indoor Environmental Quality (IEQ) causes environmental stress, which can lead to adverse health outcomes and reduced productivity. First step to enable automatic environmental control is to recognise environmental conditions that can negatively influence each individual. To this end, we developed (1) multi-sensor IEQ monitoring system to measure objectively environmental quality; (2) mobile application to collect subjective evaluation of productivity, stress and IEQ data; (3) machine learning method to use IEQ data to distinguish between positive and negative self-reports of test subjects. Experimental results with real life data, collected in four classrooms of Finnish elementary school during 18 weeks, show that IEQ sensor data allows to classify with fairly high accuracy perceptions of teachers regarding their work productivity (91%), stress (81%) and IEQ (92%). This result was achieved in person-specific training (i.e., model of each individual was trained using only his/her data), whereas accuracy of leave-one-person-out approach was notably lower. These results suggest that perception is personal and some individuals are more sensitive to environmental stressors.
AB - Considering that urban people spend majority of their time indoors, buildings should support health and productivity. However, quite commonly unsatisfactory Indoor Environmental Quality (IEQ) causes environmental stress, which can lead to adverse health outcomes and reduced productivity. First step to enable automatic environmental control is to recognise environmental conditions that can negatively influence each individual. To this end, we developed (1) multi-sensor IEQ monitoring system to measure objectively environmental quality; (2) mobile application to collect subjective evaluation of productivity, stress and IEQ data; (3) machine learning method to use IEQ data to distinguish between positive and negative self-reports of test subjects. Experimental results with real life data, collected in four classrooms of Finnish elementary school during 18 weeks, show that IEQ sensor data allows to classify with fairly high accuracy perceptions of teachers regarding their work productivity (91%), stress (81%) and IEQ (92%). This result was achieved in person-specific training (i.e., model of each individual was trained using only his/her data), whereas accuracy of leave-one-person-out approach was notably lower. These results suggest that perception is personal and some individuals are more sensitive to environmental stressors.
KW - indoor environment quality
KW - Machine learning
KW - Classification
KW - Productivity
KW - Stress
KW - self-report
UR - http://www.scopus.com/inward/record.url?scp=85076286804&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-34255-5_3
DO - 10.1007/978-3-030-34255-5_3
M3 - Conference article in proceedings
SN - 978-3-030-34254-8
T3 - Lecture Notes in Computer Science
SP - 27
EP - 40
BT - Ambient Intelligence - 15th European Conference, AmI 2019, Proceedings
A2 - Chatzigiannakis, Ioannis
A2 - De Ruyter, Boris
A2 - Mavrommati, Irene
PB - Springer
T2 - European Conference on Ambient Intelligence, Aml 2019
Y2 - 12 November 2019 through 15 November 2019
ER -