With the advancement of smart sensing technologies, embedded systems, open source OS, wireless networks, and IoT gateways, elders will have better quality living environment with self-monitoring health care. The solution we proposed will help monitor aging family members and eliminate the concerns of safety from family members who have a vested interest in their aged loved ones. An open interoperable IoT based platform architecture is a good choice for family members, relatives, and caregivers to remotely monitor elder’s sensation data in real time, such as monitoring elders’ living environmental quality and vital body sign in real time. Our technological platform prototype will involve pulse rate sensor, indoor environmental temperature and humidity sensor, Arduino and Raspberries Pi embedded systems, Firebase from Google, firebase real-time database, and web App. The developed prototype can collect individual sensational data, store and sync data among all registered users in real time. The experiment has been setup to test the IoT based platform with two sets of embedded computers with the biosensors installed at different locations. Multiple users, including local users and remote users, can concurrently receive the sensational data via our developed web in real time. The information received can inform people about the health and living conditions of the elders who live at the home independently.

Keywords: IoT, arduino, raspberry pi, data acquisition, biosensors, Google firebase, web application

The world’s older population continues to grow at an unprecedented rate. According to World Population Prospects 2019¹ one in six people in the world will be over the age of 65. The number of people above age 80 years is growing even faster than the number above age 65. In 1990 there were just 54 million people aged 80 or over in the world, a number that nearly tripled to 143 million in 2019. Globally, the number of persons aged 80 or over is projected to nearly triple again to 426 million in 2050 and to increase further to 881 million in 2100. Many elderly citizens live in nursing homes or have personal nurses to help them with daily tasks, but for those who are used to be independence, such restraint of freedom can be devastating. It is crucial to develop modest support systems for the elder to live healthily and safely in an independent living environment. With the advancement of smart sensing technologies, embedded systems, wireless communication, and IoT-enable service, elders will have enhanced quality of their living environment in self-monitoring or remote family-monitoring of health care. The research in this area is in wide range from low-level data acquisition by sensors to high-level data integration,^2‒4 including sensing and raw-data acquisition, the communication hardware and software to relay data locally or remotely, and data analysis and learning/reasoning methods to identify activities and provide caregivers and experts with useful and significant information. The different sensing technology and communication networks have been proposed.^5‒7 For example; wearable sensors can be used to acquire data to generate activity patterns for the forgotten complex activities in smart elderly living environment. Wireless and IoT (Internet of things) service can be used to send information remotely to doctors and caregivers for early detection and prevention of depression and diabetes. Accelerometers and an electrocardiogram (ECG) sensor can be used to monitor heart attack. Near-field imaging floor sensor and noise recognition patterns can be used to detect falls. Electromyography (EMG) can be used to help detect neuromuscular abnormalities. A wide range of open-source hardware that includes healthcare sensors and low-cost single-board computers are readily available on the consumer market.⁸ Both Raspberry Pi and Arduino are equally popular open source hardware but using different boards. This allows us to improve and contribute to the hardware and software design under an open-source license. An IoT driven health monitoring system using Raspberry Pi was proposed.⁹ The Raspberry Pi (RPi) connects to the sensors of heartbeat, temperature, and accelerometer via GPIO, I2C bus, and extra ADC (analog to digital converter) since there is no analog channel available from the RPi. Arduino is a very low-cost microcontroller. There are a number of libraries available in the Arduino community to enable the developer easily to access various sensors and modules that are reacting to external digital and analog signals. The hardware of ESP 8266 with Arduino board is considered.¹⁰ The analog reading from Arduino is shared to the web database via ESP 8266 Wi-Fi Module that can be accessed by the patients or registered doctors. Our previous work combined the advantages of Arduino and RPi,^11,12 in which both hardware development boards are chosen to acquire and relay the sensational data to the Google Firebase. Our system is constructed around Firebase from Google, which provide application development platform as Backend-as-a-Service (Baas). We conducted the experiment with two sets of the hardware. Each set of hardware includes Arduino microcontrollers, RPi minicomputers, and the sensors. Each set of hardware is used to send a group of acquired data to real time database via Wi-Fi. Multiple clients, including local and remote users, can concurrently receive two groups of sensational data via our developed web in real time. The information received can inform people about the health and living conditions of the elders who live at the home independently.

Sensor principle and applications

Temperature and relative humidity (RH) measurements are often collected as part of an indoor environmental quality investigation. Most people feel comfortable when the air temperature is between 68°F (20°C) and 80°F (27°C) and the relative humidity ranges from 35% to 60%. The ASHRAE guidelines recommend 68°F to 74°F in the winter and 72°F to 80°F in the summer. The guideline also recommends relative humidity of 30 to 60 percent.¹² Poor environmental conditions will have side effect to people health. Cold indoor temperatures have been associated with increased blood pressure, asthma, and depression of elders. A very hot environment often leads to elderly heatstroke and muscle cramps. Relative humidity can affect the incidence of respiratory infections and allergies. If it is too dry, the relative humidity is below 30%. The likelihood of the spread of cold and flu increases. On the contrary, the likelihood of fungi and bacteria grows quickly if the relative humidity is higher than 60%. The heart rate is a vital body sign. It is important to identify if the elders have normal range of the heart beats. The heart rate may be affected due physical exercise, sleep, anxiety, stress, and illness. An average heart rate is from 60 to 100 beats per minute (BPM). Arrhythmia can cause heat to beat slower than 60 BPM (Bradycardia) or faster than 100 BPM (Tachycardia). The DHT22/AM2303 and pulse sensors (Figure 1) are used in our experiment setup. The DHT22/AM2302 is a temperature and humidity sensor¹³ used to monitor the living environment. The sensor is made of two parts, a capacitive humidity sensor and a thermistor. It is capable of measuring 0–100% relative humidity with 2-5% accuracy and -40 to 80°C temperature readings ±0.5°C accuracy. It is simple to use, but requires two seconds sensing period to grab the data. The sensor’s output signal can use a customized digital format which can be connected to the digital input pin of Arduino Nano. The pulse sensor¹⁴ can obtain heart beating data from the users by placing the sensor on their fingertips. When a heart beats, the blood flow makes absorption of the wavelength of green light. The reflected green light will be sensed by the light photo sensor APDS-9008. When blood pumps through tissues, the converted voltage output will change due to green light absorption. The converted voltage signal then is connected to an analog input channel of Arduino to match the acquired raw data.

Figure 1 DHT22/AM2302 and pulse sensors.

The IoT enabled cross platform design

Our current research makes use of early IoT prosthetic framework. In our improved IoT enabled cross-connected platform for smart living environment (Figure 2), multiple smart objects are connected to Firebase. Arduino Nano is a very low cost microcontroller with a crystal oscillator of frequency 16 MHz and is not running an operating system. There are a number of libraries available in the Arduino community to enable the developer easily to access various sensors and modules that are reacting to external digital and analog inputs. With a Nano board development platform, the overall size of the prototype is greatly reduced. The Arduino Nano is approximately 85% smaller than the size of the Arduino Mega. Raspberry Pi 3 Model B+ could reach 1.4 GHz and Pi 4 can reach 1.5 GHz on the CPU with low cost. They are faster than Arduino by almost 90 times in clock speed. Both RPi 3 Model B and RPi 4 are 64 bits single-board computers built-in with Wi-Fi and Bluetooth features. They can serve miniature Linux servers to handle things like user interface, application, data processing, network connections, web servers, etc. Our approach is to directly connect the Arduino Nano board to the Raspberry Pi via serial communication USB cable. Since the low-cost Arduino Nano easily reacts almost immediately to a raw-data acquisition at its digital or analog channels, we connect sensors to Arduino Nano and then signal the Pi about the sensing data changes. The acquired data will be processed and relayed by the miniature computer Raspberry Pi to Firebase via Wi-Fi network connection. Our system is constructed around Firebase from Google, which provide application development platform as Backend-as-a-Service (Baas). Firebase can provide developers with a variety of tools and services, such as real time database to store data and website hosting tools to develop quality applications. We can collect and store data in Firebase real-time database as JSON (JavaScript Object Notation) objects. The data stored as JSON is synchronized to all connected clients in real-time. Our developed frontend hosting applications will retrieve stored sensational data from Firebase real-time database to allow relevant parties (doctors, family members, etc) to monitor the living conditions in real time from a mobile device or from a website.

Figure 2 IoT enable cross-connected platform for smart living environment.

Implementation, test, and analysis

Raspberry Pi does not have a standalone module for the converter from analog signal to digital signal. It only can accept the digital signal at its GPIO pins. Therefore, we choose the Arduino Nano for sensor digital or analog data acquisition. The sensed data is transmitted to Raspberry Pi from Arduino through the serial communication. The sensor DHT22/AM2303 is used to acquire the sensational data of temperature in Celsius (oC) and Fahrenheit (oF), and relative humidity (RH). The heat index is computed based on relative humidity and temperature. It referred to as the temperature felt by the body. It can be shown as heat index in (oC) and Fahrenheit (oF). For example, if temperature is 19oC/66.2oF and relative humidity RH is 31.4%, the heat index is 18oC/64oF. The body feels colder than 19oC/66.2oF. With the temperature of 28oC/82.4oF and humidity RH of 40%, the heat index goes 28oC/82oF. However, if the RH goes 80% with the same temperature of 28oC/82.4oF, the heat index will increase to 32oC/90oF.¹⁶ The body feels very hot and uncomfortable. It is a warning sign for elders to stay in safe to avoid heat exhaustion even the temperature itself is only 28oC/82.4oF. In our experiment, the data pin of DHT22 is connected to Arduino Nano digital pin 2. The 5V DC power battery or adjustable breadboard DC 5V power supply is provided to VCC pin of DHT22. Place a 10 KΩ resistor between VCC and the data pin to act as a medium-strength pull up on the data line. The “DHT. h” for DHT22 is download to be included in Arduino library. The methods of the DHT object can get the values of humidity, temperature, and heat index via read Humidity, read Temperature, and compute Heat Index, respectively. To use a pulse sensor to measure the heart beats per minute (BPM), the data line of pulse sensor is connected to the analog channel A3. The VCC line is connected to DC 5V power supply. We follow the same procedure used for the installation library to use the DHT22 sensor. The “Pulse Sensor Playground h” is downloaded to include to the Arduino library directory. After a declaration of the instance of Pulse Sensor Playground, the methods of get Beats per Minute and saw Start Of Beat can be accessed to acquire the data of the latest beats-per-minute (BPM) and the detected status of heartbeat pulse, respectively. All acquired data with associated strings then can be sent to Arduino Nano serial port at a baud rate of 9600. The USB cable is used to transmit the acquired sensational data from Arduino Nano to Raspberry Pi (RPi). The schematic of data acquisition design is shown in Figure 3. The raw data acquisition test using Arduino Nano is shown in Figure 4 where the two groups of data in BPM, Relative Humidity, Temperature, and Heat index are obtained in the interval at least 3 seconds to satisfy DHT22 sensing time. The differences of the two group data in relative humidity, temperature in Celsius, heat index in Celsius are 31.40%-31.30%=0.10%, 19.00°C-19.10°C=-0.10°C, and 17.78°C-17.80°C=-0.10°C, respectively. They fall in the accuracy range 2-5% of relative humidity and ±0.5°C accuracy in temperature readings. Both RPi 3 model B and RPi 4 are single-board computers built-in with Wi-Fi and Bluetooth features. We connect an Arduino Nano to the USB port of RPi. The Python serial module encapsulates the access for the serial port. The imported serial module provides backbends of RPi for Python running through the serial communication. The serial communication with the baud rate 9600 can be established between Arduino programming function Serial begin (9600) and Python method serial. Serial (‘/dev/ttyUSB0’, 9600). RPi receives the sensational data by calling Python code serial. Redline and stores the data in key/value pairs in the Dictionary data structure. Two sets of microcontrollers and RPi minicomputers with the sensors are setup to run the system experiment (Figure 5). One set uses RPi 4 with Arduino Nao and sensors of DHT22 and pulse, another set to use RPi 3B model instead of RPi 4. Both RPi computers are running at the same time to relay the data in key/value pairs to Firebase real-time database via the Wi-Fi mobile hotspot. Once group of pairs of key and value (key, value) are stored at the Dictionary. They are (Relative Humility, 32.60%), (Temperature in Celsius, 22.90*C), (Heat index in Celsius, 22.10*C), (Temperature in Fahrenheit, 73.22*F), (Heat index in Fahrenheit, 71.77*F), and (BPM, 77). All the values are in real-time. The key-value pairs will be uploaded to a Firebase database via Wi-Fi and then distributed to the chosen endpoints.

Figure 3 Schematic of data acquisition system.

Figure 4 Raw data acquisition test using Arduino Nano.

Figure 5 Experiment setup with two set of hardware and sample key-value pairs.

Access to firebase

Firebase is built on Google infrastructure and is a cloud-based service. Backend as a Service (BaaS) is currently used in Firebase. BaaS provides the backend for mobile and web applications to integrate with their application backend. A server-side rule was created when register. The rule adds the user object to the JavaScript Object Notation (JSON) database. Only authenticated object monitors can read or write data to interact with the database. The sensor and rule models are generic objects. The database can store health and environmental data as JSON tree. Data stored as JSON is synchronized to all connected clients in real-time. Firebase provides tools to allow the access of the real time database by a number of programming languages, including python and JavaScript. A python program running on RPi is used to receive the data from Arduino platform and transmit them via WiFi to Firebase. After the Firebase project is created with service account, we can install Firebase Python SDK and import fire base_ admin to read and write real-time database data with full admin privileges, or limited privileges. Each RPi have unique CPU serial number that can be obtained by imported function open('/proc/cpuinfo','r'). When fetching the CPU serial number at the location in the database, we can retrieve all of its associated child nodes. Even though a JSON data tree can support thirty-two levels deep in each path, in practice, we opt to keep the data structure as flat as possible. This allows faster data access. Figure 6 shows JSON objects from the Firebase real-time database. The CUP serial number RPi3_id=0000000035765534 is for RPi 3 model B and RPi4_id=1000000047400fe0 is for RPi 4. When fetching RPi3_id, it can retrieve all child nodes where the values are acquired by its set of sensors. RPi3_id child nodes show that owner is Test 2 in Jan 2020 and the key-value are: (BPM, 77), (Heat index in celsius, 22.10*C), (Heat index in Fahrenheit, 71.77*F), (Relative Humility, 32.60%), (Temperature in celsius, 22.90*C), and (Temperature in Fahrenheit, 73.22*F). When fetching RPi4_id, it can retrieve all child nodes where the values associated with its sensational data. RPi4_id child nodes show owner Test 1 in December 2019 and the key-value: (BPM, 72), (Heat index in Celsius, 19.91*C), (Heat index in Fahrenheit, 67.83*F), (Relative Humility, 28.80%), (Temperature in Celsius, 21.00*C), and (Temperature in Fahrenheit, 69.80*F). Frontend hosting applications to retrieve JSON nodes from the Firebase real time database allows relevant parties (doctors, family members, etc) to monitor the living environment and health in real time from a mobile device or from a website. A website can be constructed using tools provided by Firebase to build html/JavaScript file to display the stored data in a real-time database. An example of the webpage that displays sensor data updated in real time, in synchronization with the real-time database and the sensor data received by the python program running on Raspberry Pi, are shown in Figure 7.

Figure 6 JSON object from Firebase real-time database.

In Figure 7, the real-time database with JSON nodes is shown at the left side of the figure. The webpage, with URL in firebaseapp.com domain at the right side, is from a web browser on a PC computer. They both displays the five data values stored in Firebase database for each Tester (Client): The heart beat per minute (BPM), the humidity (Humidity RH), the temperature (Temperature in *C), the heat index (Heat index in *C), the temperature (Temperature in *F), and the heat index (Heat index in *F). The Raspberry Pi runs a Firebase web server so it shows the same two groups of user data values (top right side of the Figure 7) stored in Firebase database. Therefore, when the python program that receives the sensor data from Arduino is running, the changes of sensor data values in real time are captured in the browser DevTools console. Data stored as JSON is synchronized to all connected clients in real-time. An authorization object monitors and manages the ability of users to interact with the database. With the communication hardware and software, the sensing data can be sent locally or remotely to provide caregivers and experts with useful and significant information.

Figure 7 Sync data with NoSQL cloud database across all of the clients in real time.

The project deployed IoT enabled cross-connected platform for smart living environment. The multiple platforms communicate each other to send sensing data via serial communication to Raspberry Pi from Arduino Nano, then have the Raspberry pi Wi-Fi relay sensing data to the Firebase real-time database, which is accessible through our developed website. The cloud-based solution has no restrictions to geographical location. The real-time information can be monitored locally or remotely. The system has multiple platforms to satisfy user needs. However, we found out the pulse rate sensor has not been working properly such as inconsistent BPM readings. Our further work will include calibrating the sensational information to remove the noises, develop mobile Arduino app with low energy Bluetooth connection, add more user-friendly vital sign sensors into the platform, such as non-contact infrared temperature sensor to acquire body temperature, and apply learning model to forecast and predict health conditions.

This research is supported by PSC-CUNY Research Awards 61463-0049. We would like to thank CET (Computer Engineering Technology) graduated student Mohamed Alborati.

Authors declare that there is no conflict of interest.

1. World Population Prospects 2019: Highlights (ST/ESA/SER.A/423).
2. Qin Ni, Ana Belén García Hernando, Iván Pau de la Cruz. The Elderly’s Independent Living in Smart Homes: A Characterization of Activities and Sensing Infrastructure Survey to Facilitate Services Development. Sensors (Basel). 2015;15(5):11312–11362.
3. Patel S, Park H, Bonato P, et al. A review of wearable sensors and systems with application in rehabilitation. J NeuroEngineering Rehabil. 2012;9:21.
4. Laplante PA, Kassab M, Laplante NL, et al. Building Caring Healthcare Systems in the Internet of Things. IEEE Systems Journal. 2018;12(3):3030–3037.
5. Uddin MZ, Khaksar W, Torresen J. Ambient Sensors for Elderly Care and Independent Living: A Survey. Sensors (Basel). 2018;18(7):2027.
6. Ni Q, García Hernando AB, de la Cruz IP. The Elderly's Independent Living in Smart Homes: A Characterization of Activities and Sensing Infrastructure Survey to Facilitate Services Development. Sensors (Basel). 2015;15(5):11312–11362.
7. Yu Wang, Sunghoon Jang. A pulse sensor interface design for FPGA based multisensor health monitoring platform. International Journal of Biosensors & Bioelectronics. 2019;23-27.
8. Cippitelli E, Fioranelli F, Gambi E, et al. Radar and RGB-Depth Sensors for Fall Detection: A Review. IEEE Sensors Journal. 2017;17(12):3585–3604.
9. Niezen G, Eslambolchilar P, Thimbleby H. Open-source hardware for medical devices. BMJ Innov. 2016;2(2):78–83.
10. Kamal N, Ghosal P. Three Tier Architecture for IoT Driven Health Monitoring System Using Raspberry Pi. 2018 IEEE International Symposium on Smart Electronic Systems (iSES) (Formerly iNiS). Hyderabad, India. 2018. pp. 167–170.
11. Rahman RA, Aziz NSA, Kassim M, et al. IoT-based personal health care monitoring device for diabetic patients. IEEE Symposium on Computer Applications & Industrial Electronics (ISCAIE). Langkawi. 2017. pp. 168–173.
12. Yu Wang, Hunter W, Chen X, et al. Improved Hardware Design of IoT Prosthetic Device. The 2018 ASEE Mid-Atlantic Fall Conference. Brooklyn, NY. 2018.
13. Martinez G, Chen X, Lai T, et al. IoT in Myo-Prosthetics. 2018 ASEE Northeast Section Annual Conference. Hartford, CT. 2018
14. https://learn.adafruit.com/dht
15. https://www.broadcom.com
16. https://www.wpc.ncep.noaa.gov/html/heatindex.shtml