In this paper, a tool for the simulation of home environment capable of representing human moving patterns, related to Activities of Daily Living (ADL), and modelling Passive Infrared (PIR) sensor networks is presented. PIR sensors have been chosen because they are non-intrusive, contactless and low-cost. The tool has been programmed in MATLAB and provides a graphic interface from which the developer can change key simulation parameters. It makes it possible to load and visualise the 2D map of the home environment used for simulation as well as add, customise and place ceiling or wall mounted PIR sensors, and regulate users' trajectory parameters, such as walking speed, step length or path efficiency. The simulator is useful for quickly generating synthetic data to train machine learning (ML) algorithms able to recognize user behavior, without the necessity to perform long acquisition periods. In order to demonstrate its applicability, the tool has been used to create normal and wandering trajectories and their related sensor activations. These data were employed to develop a ML algorithm able to detect overnight wandering, a common behaviour in patients with dementia. The results show that a Decision Tree (DT) algorithm is reliable for the purpose of distinguishing normal trajectories from the wandering ones detected by PIR sensor activations, obtaining an accuracy level of over 95% using a cross-validation approach.

Measurement of activities of daily living: A simulation tool for the optimisation of a passive infrared sensor network in a smart home environment

Casaccia S.;Rosati R.;Scalise L.;Revel G. M.
2020

Abstract

In this paper, a tool for the simulation of home environment capable of representing human moving patterns, related to Activities of Daily Living (ADL), and modelling Passive Infrared (PIR) sensor networks is presented. PIR sensors have been chosen because they are non-intrusive, contactless and low-cost. The tool has been programmed in MATLAB and provides a graphic interface from which the developer can change key simulation parameters. It makes it possible to load and visualise the 2D map of the home environment used for simulation as well as add, customise and place ceiling or wall mounted PIR sensors, and regulate users' trajectory parameters, such as walking speed, step length or path efficiency. The simulator is useful for quickly generating synthetic data to train machine learning (ML) algorithms able to recognize user behavior, without the necessity to perform long acquisition periods. In order to demonstrate its applicability, the tool has been used to create normal and wandering trajectories and their related sensor activations. These data were employed to develop a ML algorithm able to detect overnight wandering, a common behaviour in patients with dementia. The results show that a Decision Tree (DT) algorithm is reliable for the purpose of distinguishing normal trajectories from the wandering ones detected by PIR sensor activations, obtaining an accuracy level of over 95% using a cross-validation approach.
978-1-7281-4460-3
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11566/283960
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 3
  • ???jsp.display-item.citation.isi??? ND
social impact