Table 8.

Summary of research works that use multiple types of ambient sensor technology.

Research Authors (Year)	Purpose	Characteristics	Outcomes	Sensors
Alwan et al. [118] (2006)	Recognition of activities of daily living	Systems for detecting activities of daily living were installed in 15 assisted living units. The reports were sent to professional caregivers of the residents. Fifteen residents and six caregivers participated in the system. It was a pilot study in which monitoring was performed for three months. Quality of life was assessed using a standard satisfaction-with-life scale instrument.	There was a high acceptance rate of the system. The approach could be used for improved healthcare planning and detection of health status changes.	PIR motion sensors, stove sensor, bed pressure sensor.
Alwan et al. [119] (2006)	Recognition of activities of daily living	Activities of daily living were monitored for 26 elderly residents and 25 caregivers over four months. A standard satisfaction-with-life scale instrument was used to assess the quality of life of the elderly people and the caregivers.	Once four months of monitoring were finished, there was no significant difference in the quality-of-life scores of the elderly users and the caregivers. The system seemed to be highly acceptable.	PIR motion sensors, stove sensor, bed pressure sensor.
Alwan et al. [120] (2007)	Recognition of activities of daily living	The purpose of the work was to assess the impact of passive health status in assisted living. Two aspects were analyzed: the cost of care and the efficiencies of caregivers. Activities of daily living systems were monitored for 21 residents for over three months.	The study demonstrated that the monitoring technologies that were used in the work significantly reduced billable interventions, hospital days, and cost of care to players. Moreover, they had a positive impact on professional caregivers’ efficiency.	PIR motion sensors, stove sensor, pressure sensors.
Ariane et al. [121] (2012)	Fall detection	The proposed fall detection system was simulated by testing on scenarios in an existing data set.	89.33% accuracy.	PIR motion sensors, pressure mats.
Bemis et al. [122] (2008)	Recognition of activities of daily living	It was a case study on two residences based on seven and four months of monitoring.	The functionality of the system in detecting activities and deviations in patterns of activities was described.	Video monitoring, PIR motion sensors.
Bemis et al. [123] (2010)	Recognition of activities of daily living	The work reported the progress in sensors, middleware, and behavior interpretation mechanisms, spanning from simple rule-based alerts to algorithms for extracting the temporal routines of the users.	The functionality of the system was demonstrated.	Video monitoring, PIR motion sensors.
Celler et al. [124] (1996)	Recognition of activities of daily living	The work presented a smart home monitoring system that was based on sequences of pressure. It mainly focused on pressure transfers in the bedroom and bathroom to check whether the motion evaluation is in the normal range or not.	The functionality of the system was demonstrated. The system showed encouraging results for precise fine-grained activity monitoring systems, especially using high-precision user localization sensors.	PIR motion sensors, sound sensors, temperature sensors, light sensors, pressure sensors.
Chung et al. [125] (2017)	Sleep stage classification	A novel approach was proposed for sleep stage classification using a doppler radar and a microphone. The classification algorithm was designed based on a standard polysomnography reference-based database and medical knowledge of doctors and sleep technologists at a hospital. The algorithm outperformed commercially available products for a specific database.	100% accuracy.	Doppler radar and microphone.
Guettari et al. [126] (2010)	Localization	This work proposed a localization system that was based on a combination of infrared sensors and sound sensors. The system mainly used the azimuth angles of the sources. This multimodal system improved the precision of localization compared to a standalone system.	54% improvement was achieved using the proposed multimodal system compared to a standalone one.	PIR motion sensors and sound sensors.
Kinney et al. [127] (2004)	Recognition of activities of daily living	It was a pilot study on 19 families for activity monitoring. Monitoring was performed for six months.	The main advantage of the system was the ease of tracking the users. The main disadvantage was the annoyance that was created by false alerts. The cost was $400 to equip the home. Ninety dollars per month was the cost of maintenance.	Video camera, PIR motion sensors.
Lotfi et al. [128] (2011)	Recognition of activities of daily living	It was a case study on two dementia patients. The first patient was monitored for 20 days. The second patient was monitored for 18 months.	The system was used to identify abnormal behavior. The system demonstrated satisfactory performance in identifying health status using different ambient sensors.	PIR motion sensors, door opening sensors, flood sensors.
Rantz et al. [129] (2008)	Fall detection	A case study was performed for retrospective analysis of fall detection data.	A change of health status was detected by the system but ignored by the nurses.	Video camera, PIR motion sensors, bed pressure sensors, door sensors.
Van Hoof et al. [130] (2011)	Recognition of activities of daily living	It was a pilot study for daily activity monitoring and fire wandering detection. The system was installed in the range of 8–23 months for analysis.	Use of the proposed system improved the sense of safety and security.	PIR motion sensors, video camera.
Zhou et al. [131] (2011)	Recognition of activities of daily living	The work tried to recognize simulated activities that were monitored in testbed for a month.	92% precision; 92% recall.	Video camera, PIR motion sensors.
Zouba et al. [132] (2009)	Recognition of activities of daily living	The authors recognized simulated activities that were monitored in a laboratory setting.	62–94% precision; 62–87% sensitivity.	Video camera, PIR motion sensors.
Zouba et al. [133] (2009)	Recognition of activities of daily living	The work was focused on monitoring simulated activities in a laboratory setting.	50–80% precision; 66–100% sensitivity.	Video camera, PIR motion sensors.