Introduction: Newborns Monitoring System (Based on Wearable Temperature Sensors and Intel's Edison)

The problem

“A major cause of neonatal deaths worldwide is infection (36% of death cases, which includes sepsis/pneumonia, tetanus and diarrhea)… fever is important as it may be the first and only sign of a serious infection. If temperature is just slightly above the normal range the newborn baby must be brought to the doctor as soon as possible…The normal range for a temperature taken under the armpit is 36.7°C to 37.1°C” [1]

“The majority - almost 3 million of these die within one week and almost 2 million on their first day of life….Virtually all (99%) newborn deaths occur in low- and middle-income countries. It is especially in Africa and South Asia….Almost 3 million of all the babies who die each year can be saved with low-tech, low-cost care.” [1]

What can be done?
“For infections, antibiotics is life-saving and needs to be available locally. Low birth weight babies need to maintain body temperature through skin-to-skin contact with the mother. Several of the above interventions would also help save the lives of mothers and prevent stillbirths” [2]

Further reading :


What’s our solution?

We aim to provide a low-cost early detection system for underdeveloped countries and remote areas, suffering from high newborns mortality, with lower internet access. Out system is comprised of wearable temperature sensors connected to an Intel Edison board which monitors infants (and mothers in a later version) temperatures in a hospital room (or indoors). If the temperature exceeds -or falls behind- the normal range, even slightly, instant warnings trigger, notifying nurses. The warnings are sonar via a buzzer and visual via an RGB LCD encoding the infant’s state in color intensities. If the temperature goes out of the critical range the closest doctor will be notified via sms and email.

Data is permanently sent via WiFi to another Intel Edison board (Server) to be stored in a local database, to keep track of the each infant’s temperature that could be accessed by nurses or doctors to check if there’s any abnormalities. A local web application allows nurses to update the infant’s information, or create new infants’ entries.

For later data mining, the database’s content could be sent to cloud services for analytics, to detect tendencies and also to map areas with high infants infections, so more care can be provided where it’s needed the most.

Step 1: Set Up the Edison

To carry out the project, Intel® Edison was used with the Intel® Edison Board for Arduino.

We use a Grove Base Shield along with a couple of Grove Temperature Sensors, Grove - Buzzer module

and with a Grove LCD - RGB Backlight.

So, let's go back to the Edison setup. Follow these steps:

Assembling the Intel® Edison board with the Arduino*expansion board

Step 2: Connect the Base Shield With the Sensors and the LCD

This step is pretty simple as you only have to plug in the temperature sensors, the buzzer and the LCD display to the base shield, as it is showed in the wikis linked in step 1.

Step 3: Start Monitoring the Newborns Temperature [1]

As you can see the video,

Our system is comprised of wearable temperature sensors connected to an Intel Edison board which monitors infants (and mothers in a later version) temperatures in a hospital room (or indoors). If the temperature exceeds -or falls behind- the normal range, even slightly, instant warnings trigger, notifying nurses in the hospital. The warnings are sonar via a buzzer and visual via an RGB LCD encoding the infant’s state in color intensities. If the temperature goes out of the critical range the closest doctor will be notified via sms and email.

Step 4: Start Monitoring the Newborns Temperature [2]

For the sake of the test, we used dummy values to show the system's reactivity. As you can see in the photos, the screen is blank green when everything is fine and newborns temperatures are normal, but once an abnormality is detected the buzzer with stigger an alarm soud for the nurses, and the LCD encodes each ne

Step 5: Sending Monitoring the Data to Node.JS Server on Another Local Edison

Intel® IoT Invitational

Participated in the
Intel® IoT Invitational