Jaimico: The new e-Health Care Companion Robot

Story

The so-called Internet of Things (IoT) can be seen as a decentralized, loosely coupled, and dynamic system in which a large number (perhaps billions) of everyday objects are endowed with interconnection capabilities to augment their own capabilities and to be able to communicate and cooperate with other devices. The idea of IoT is to make this interconnection possible despite the great physical and functional heterogeneity of these objects or devices. In this way, objects become special participants in any kind of industrial, logistic, domotic, social, health, etc. process.

One of the fields of use of these IoT devices is in medicine, creating a new field called IoMT or Internet of Medical Things. IoMT ranges from medical devices found in hospitals and clinics, to small devices that capture biosignals from users. One of the most interesting aspects of IoMT devices is their ability to connect to the network. This allows the mass sending of biosignals such as an electrocardiogram (ECG), blood sugar level, blood pressure, photoplethysmography (PPG), SPO2 levels, etc. These signals are sent to large servers, which store and analyze this information. Companies such as Amazon offer these data storage, pre-processing, and processing services. Allowing companies and industries to create applications that help people's well-being.

The project presented below is a system that is part of the IoMT and uses Amazon AWS services for the collection of data signals. AWS provides a secure data transfer channel between devices and its services. Channels such as AWS IoT, which enable fast, secure, and scalable communication between any IoT device and AWS.

This is "Jaimico"

My robot "Jaimico" is an integration of different devices and AWS services. Jaimico is a personal assistant for daily use anytime and anywhere. Jaimico incorporates the Alexa voice assistant created by Amazon, this assistant is embedded inside an M5Core2-AWS.

One of the ideas that Jaimico shares are the decentralization of the concessions, this means that each device is completely autonomous and able to communicate with AWS without having to pass the data through the robot. In this way, the acquisition of heart signals (ECG) is not performed by the robot, but this does not mean that the robot cannot control this acquisition task.

This decentralization means that the system can even be used by third parties, i.e. if Jaimico is monitoring an elderly person and his caregiver or child wants to know how he is doing. They can access from any terminal with Alexa and activate the Jaimico skill and ask it to perform the analysis. Each of the systems sends the respective signals to AWS and stores them in a DynamoDB database. The storage of these signals is done with the date and time in which it was acquired, allowing to have a history in the time of all signals. Thus, if something happens to this person, the doctor could use this history to try to determine the trigger that led to this failure. This is of vital importance because when a person goes to the doctor, he takes a sample during the consultation period and he does not know if during this time the person is suffering from this failure.

Figure 1 shows what Jaimico looks like, which was designed as a companion robot. It can be put on the shoulder. Figure 1 shows what Jaimico looks like, which was designed as a companion robot. It can be placed on the shoulder, this location makes it ideal for interaction processes.

Jaimico allows his body to be upgraded using Lego parts, which means that his appearance can change every day. This is ideal if Jaimico is used for monitoring children, as they can customize their robot as they wish.

Bio-Signal Capture Devices

Jaimico controls 4 bio-signal capture devices. The first is a chest harness, which captures the heart signal. To perform this capture, the Cardio-Chip created by Neura Sky BMD-101 is used. This Cardio-Chip was specially designed to capture ECG signals for wearable devices. It communicates through serial communication, which makes it ideal for this type of device. The BMD-101 interfaces with an M5StickC-Plus, which receives the data sent by the BMD-101. The M5StickC-Plus preprocesses this data and encapsulates it to be sent to the AWS. At the same time, the M5StickC-Plus can plot the acquired ECG signal using its LCD display.

The second device developed, the I-Health, is an ECG, muscle activity, and electrical skin activity or EDA signal capture station. It was designed using an M5EPD.

EThe third device is the Bio-Ring, a ring that allows to measure skin resistance and to obtain photoplethysmography. This can be used to determine the user's stress levels, as well as determine oxygen saturation levels or SPO2.

The last device to be incorporated was the cough analysis and temperature taking system. This device integrates a tinyMl programmed e-learning model, which is able to detect whether the person has Covid-19 or not. For this, the Virufi-Covid database was used to train the model. The system captures the cough signal using the microphone of the M5Cores2, this signal is stored inside a buffer with a size of 128. Once stored, the ML model predicts whether that cough sound is from a person with Covid-19 or not.

Video