ADL monitoring of Elderly in homes using Gas sensors

Project Information

EDGE

Grant agreement ID: 861782

DOI

10.3030/861782

Project closed

EC signature date 13 September 2019

Start date 17 December 2019

End date 16 December 2020

Funded under

• INDUSTRIAL LEADERSHIP - Innovation In SMEs

Total cost

€ 86 875,00

EU contribution

€ 86 875,00

Coordinated by

N-VISION SYSTEMS AND TECHNOLOGIES SL
Spain

Periodic Reporting for period 1 - EDGE (ADL monitoring of Elderly in homes using Gas sensors)

Reporting period: 2019-12-17 to 2020-12-16

Summary of the context and overall objectives of the project

The growing ageing population is of specific concern in Europe as it noted in the latest EU ageing report which estimates that by 2060 people ageing 65+ will reach 28% of the population, compared to 18% in 2020. These forecasts bring multiple challenges for European healthcare and financial systems. While access to Long-Term Care for older adult populations is one of the social benefits that Europe takes pride in, this access is not optimal as some care services are underutilised despite the seemingly higher demand. On the other hand, a growing population across the EU is witnessing (and contributing to) a shortage of healthcare workers and the gap is seen to be widening over time. Technology solutions based on smart home alternatives can help reduce the cost of ageing in place since it is not necessary to keep a caregiver at home 24 hours a day to track the health condition of the elderly in real-time and has tremendous potential to address the needs of the European Silver Economy.

In this context, gas sensor-based systems have been successfully deployed in controlled domestic environments for human activity monitoring. These systems can detect any activity that alters the composition of the air, such as cooking, bathing, bathroom visits, sleeping, etc. and dangers such as fire, gas leak, bad ventilation, food in bad conditions, among others. Thus, a smart electronic sensing system for continuous monitoring of the air composition together with the correct identification of normal and abnormal patterns can be used for the monitoring of the activity of daily living for older people living alone, allowing not only the elderly to live independently longer but also the caregivers to remotely follow their daily habits thus enabling a timely intervention.

The overall objective of EDGE project is to combine gas sensing technologies with advanced artificial intelligence algorithms to continuously monitor and analyse ADLs and then to detect anomalous situations based on a smart analytics engine. The successful application of these systems will meet the specifications and requirements of caregivers to improve the elderly care services and will also gain elders acceptance.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The initial study of the actual market requirements and the state of the art related to EDGE project together with a preliminary analysis of available gas sensors in the market have allowed the selection of suitable gas sensors array that meet the optimal specifications in terms of sensitivity, selectivity, detection limits, response and recovery time and sensor drift. Indeed, the four Metal Oxide (MOX) gas sensors present in the array are selected to cover most of gases and volatile organic compounds present at home due to elderly activity. In addition, a temperature sensor, a relative humidity sensor and a CO2 sensor have been housed within the same array to increase the characterisation level of the studied activities. The ADL monitoring system-based gas sensors was installed at different rooms to cover sleeping, cooking, bathing, and toileting activities as depicted in figure 1. The location of the ADL monitoring system at each room has been chosen taking into account two constraints: 1) to be close enough to gases or vapours released from the activity in order to obtain a significant gas sensor response matching the undergoing activity and 2) to be far enough from the activity in order to make the user comfortable and also protect the gas sensors system from any accident (water, chocs, etc.). The results of EDGE project demonstrate the ability of the deployed gas sensors-based system coupled with the developed machine learning models to monitor not only the daily activities of older adults but also to detect some abnormal events that would have harmful effect on their health. Figure 2 shows an example of sleeping activity identification using the principal component analysis score plot and figure 3 shows the response of the response of the ADL monitoring system-based gas sensor towards 24h of kitchen activities. Nevertheless, the complexity of the environments where some ADLs is taking place such as bathroom where all personal hygiene activities are performed makes the achievement of high models’ accuracy rate very challenging. Therefore, the development of new strategies using gas sensors in conjunction with the adoption of deep learning paradigm could improve the characterisation of such ADL.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The majority of older adults want to remain at home as long as possible in normal cases, and in some critical cases, for example, what happened during the quarantine of the COVID-19; they could not even meet their doctors or families during a specific period (by fear or due to imposed restrictions) of over two months. In normal situations, families need to offer them some assisted living facilities, e.g. nursing homes, with an average cost of around 40K euros per year. Here, ICTs and Artificial Intelligence can help older people and make this task easier and cheaper. Thus, recently many companies around the world provide a professional service to families, people receiving support, subscribers, residents and policyholders. These companies are turning to home-based enabling technology such as wearable sensors (e.g. smartwatch) and assistive devices (e.g. cameras) as tools to achieve solutions on a daily basis. The wearable sensors are very efficient in such cases, however, the “always on-the-body” requirements make older adults difficult to comply with, especially at home. In turn, using cameras have proven useful in older adults’ care to follow-up their behaviour and activities. However, as their abilities expand, questions arise regarding privacy, dignity and misuse. Gas sensors are among the suitable alternatives enabling the monitoring of older adults’ daily activities while being non-wearable, unobtrusive and privacy protecting.

EDGE project was a great opportunity for the innovative associate and NVISION to develop a non-invasive, non-obtrusive, non-wearable and privacy-respectful solution to monitor ADL of older adults living alone. This solution falls at the intersection of the current COVID-19 pandemic obligations and the rise of remote/telecare services where caregivers are still looking for low-cost and efficient solutions to monitor their care-receivers and then keep them healthier and safe as much as possible.

The achievement of this objective and its successful integration into NVISION's IoT platform (enSenior) will help to improve the elderly quality of life and reduce healthcare and formal caregivers' services costs. This solution will also assist informal caregivers to keep a “close/remote” look on their loved ones.

24h response of the ADL monitoring system-based gas sensors to screen kitchen activities 30/11/20

Installation of the ADL monitoring system inside the pilot site

Scores plot of a PCA performed on the ADL monitoring system data from 00 am to 09 am