THE INSTITUTE Because senior citizens are at a higher risk of suffering serious complications from COVID-19, medical officials have cautioned them to stay at home. Their family and caregivers have also cut back on visits due to fear of spreading the virus to them. This situation has left many of the elderly increasingly isolated and without anyone to keep a watchful eye on their health.
A team led by electrical engineer Rogerio Dionisio at the Polytechnic Institute of Castelo Branco, in Portugal, has created a wearable monitoring system that keeps track of a person’s daily routine and alerts caregivers if it suspects something is amiss.
The IEEE senior member is the deputy director of the university’s School of Technology.
The Institute asked Dionisio about how the system works.
This interview has been edited and condensed for clarity.
What problem are you trying to solve?
The COVID-19 pandemic and [the need for self-isolation] has decreased the frequency of visits [to senior citizens] by family members or caregivers. Because of this, many health concerns [of the elderly] are unresolved.
We created a system called Zelar@CB that monitors the daily activities of isolated elderly people who do not have access to at-home care. Zelar@CB detects unusual activity by monitoring the person’s usage of electrical appliances to track if the user left the stove on or if there is a change in usage. The system sends an alert to family members or caregivers if it detects that. It can also detect when the user falls.
What technologies are you using?
The system uses a LoRa [low-power wide area] Network, combined with a low-power wearable device that can be worn as a bracelet or embedded in clothing, and an energy monitor that connects to the user’s electrical appliances. The wearable device is equipped with an accelerometer [which measures acceleration] and a gyroscope [which measures an objects orientation]. The gyroscope can detect if the user has fallen.
The energy monitor is connected to the main electrical line [of the home]. The system monitors the power consumption of each electrical appliance and timestamps related to the users’ daily activities. Zelar@CB also uses an artificial intelligence algorithm [that has been] previously trained with data on the user’s regular consumption of power. It may take up to two weeks [for the AI algorithm] to be fully trained and recognize how [the user] operates home appliances.
The energy monitor uses Wi-Fi to send either an SMS or email alert, or a message through the mobile application we developed. Because the amount of data produced by the energy monitor is substantial, the monitor cannot use LoRa alone. The LoRa system implements [the] Fair Access Policy: 10 downlink messages and 30 seconds uplink time on air per 24 hours, [per device]. Although this is adequate for [the rare] fall detection events, it is not enough for regular energy consumption measurements.
Explain how your project works.
When the low-power device identifies an abnormal situation, such as the user has fallen or a change in the person’s energy consumption, it creates an alert. The alert is sent to a family member or caregiver through either the mobile app or by SMS or email.
Once someone responds to the alert and helps the user, the family member or caregiver notifies the system. We are consulting the health and legal authorities to find the best way to handle a situation where [no one answers the alert] without infringing the privacy and anonymity of the senior citizen.
What challenges have you faced, and how did you overcome them?
The COVID-19 pandemic changed the way the team worked together. Our lab was closed for several months, therefore we used videoconferencing tools and improvised IoT labs at home [to complete the project].
What is the potential impact of the technology?
The technology can help reduce the response time in case of an emergency.
How close are you to the final product?
We are working with senior citizens who are living in Castelo Branco, Portugal, to conduct tests with our prototype.
How can other IEEE members get involved?
We have created a GitHub repository, called Zelar-CB, to share our code. Members can contribute to the project with new ideas, algorithms, and code optimization techniques.