Taking advantage of an existing indoor climate monitorization for measuring occupancy

Unai Saralegui, Miguel Angel Anton, Joaquin Ordieres-Mere

Research output: Contribution to journalArticlepeer-review

7 Citations (Scopus)

Abstract

This paper describes a procedure to gain additional information from an already existing infrastructure primarily designed for other purposes. The deployed sensor network consists of wirelessly communicated indoor climate monitoring sensors, for which it is tried to extend its usage by determining occupancy in the room they are located, in that way the system provides a higher level aspect of the house usage. An elderly caring institution’s building has been monitored for one year obtaining data about temperature, relative humidity and CO2 levels from five different rooms. Such data shows some interesting patterns as the air flow between the rooms which should be considered in any real case scenario. The data has been used to train some machine learning models, which show acceptable quality overall suggesting to use this kind of sensing equipment to perform an occupancy monitoring non-intrusively. The acquired knowledge could bring additional opportunities in the care of the elderly, especially for specific diseases that are usually accompanied by changes in patterns of behaviour. By using the occupancy status it could be possible to determine changes in the daily patterns in that segment of the population which could be an indicative of the initial states of a disease or a worsening in it.
Original languageEnglish
Pages (from-to)327-334
Number of pages8
JournalWSEAS Transactions on Environment and Development
Volume13
Publication statusPublished - 2017

Keywords

  • Domestic occupancy
  • Climate sensors
  • Pattern analysis
  • Health monitoring
  • Smart buildings
  • Machine learning
  • Internet of things

Fingerprint

Dive into the research topics of 'Taking advantage of an existing indoor climate monitorization for measuring occupancy'. Together they form a unique fingerprint.

Cite this