HazelBot Smart Medical Assistant
Doctors and health assistants play an important role in the hospitals and other health services. Due to the recent pandemic, the healthcare workforce has been overburdened in performing their tasks. Smart medical Assistants can reduce this workload and help the medical industry by performing specifi
2025-06-28 16:27:33 - Adil Khan
HazelBot Smart Medical Assistant
Project Area of Specialization Electrical/Electronic EngineeringProject SummaryDoctors and health assistants play an important role in the hospitals and other health services. Due to the recent pandemic, the healthcare workforce has been overburdened in performing their tasks. Smart medical Assistants can reduce this workload and help the medical industry by performing specific and trivial tasks. Hence, the idea behind the project ‘HazelBot: Smart Medical Assistant’ is to design a smart robot that can assist doctors and medical professionals in performing their daily tasks. Proposed design will use artificial Intelligence to predict health condition and interact with the patients. It will use an array of sensors interfaced with Esp32 microcontroller and placed inside a wearable hand-wrap. Hand-wrap will acquire body temperature, pulse rate, blood pressure and oxygen level of the patient. The acquired data will be transmitted wirelessly to the Raspberry Pi for further processing. Smart robot will predict patient’s health condition using machine learning and instruct the patient or the doctor through email and messages. Robot will use speech recognition and synthesis to communicate with the patient. Speaker, microphone and display will be interfaced with the Raspberry Pi for Speech and graphical interface.
Project ObjectivesMain objectives of the project are:
- A smart medical assistant that can measure different vital signs of the human such as body temperature, blood pressure, heartbeat and oxygen level using a wearable wrap
- Wireless transmission of patient's vital signs to the main controller
- Machine learning based prediction of patient health condition and intimation to health care professionals
- AI based verbal interaction with the patient using speech recognition and synthesis
- A patient’s report will be generated by the robot and sent to the doctor
Design of HazelBot: Smart Medical Assistant is composed of different Hardware components that are integrated together. Hardware components are selected in way that the overall cost and design remain simple and easy to understand. Main components of the design are discussed below:
Sensors and Data Acquisition
In order to acquire vital signs such as, temperature, oxygen level, blood pressure and heart rate different sensors have been used. All the sensors are non-invasive which means they do not prick in the body Sensors were placed inside the hand wrap which will be worn by the patient. All the sensors have been directly interfaced with the ESP32 microcontroller in order to get the values of the vital signs.
Microcontroller
The microcontroller will do two main things, firstly it will acquire data in regular intervals. Secondly, it will send this data regularly towards the server. ESP32 microcontroller has been used to acquire data from the sensors. Acquired data will be wirelessly transmitted to the main controller using WiFi and MQTT protocols.
Main Processor
Raspberry Pi 4 has been used to receive data from ESP32 wirelessly. All the acquired data from the sensors is sent to the Raspberry Pi, hence it is placed inside the robot body. Machine learning algorithms will be processed on the Raspberry Pi board as well as speech processing. It is interfaced with external devices such as the LCD, mic and the speakers, in order to communicate with the patients and the doctor.
Machine Learning Based DiseasePrediction
Machine learning algorithms are used, in order to predict the patient’s health after receiving the data from the ESP32 microcontroller. To train the machine learning model data sets will be collected from Kaggle and other sources. Machine learning models that will be explored include Support Vector Machines (SVM), Random Forests and Neural Networks. Trained machine learning algorithm will predict patient disease and share this information with the health care professionals. The disease predicted are:
- COVID-19
- Malaria
- Arrythmia
- Asthma
- Hypertension
Artificial Intelligence Based User Interaction
Speech recognition and synthesis have been used in order to recognize the vocals of the patients. Then the system will process it and give an intelligent answer to the patient as a response. This will allow successful communication between the HazelBot and the patients. A mic, speakers and an LCD screen will be interfaced with Raspberry Pi to make the user interaction more effective and user friendly.
Project block diagram is shown below:
The key benefits of proposed project are following:
- HazelBot will reduce the workload of doctors and the medical staff. This will be beneficial in our public hospitals where medical staff is always overburdened
- HazelBot will perform specific medical tasks and will reduce the risk of exposure to pandemics like Covid-19 for medical staff
- HazelBot will measure multiple vital signs of the patients which will be beneficial in diagnosing early stages of many diseases.
- Medical assistants like HazelBot will be active all the time and can efficiently work for long hours
- HazelBot can be used in educational institutes, shopping malls and many other public places for medical assistance
The technical details and final deliverable of the project are given as follows:
- Sensor Interface with Esp32 and Data Acquisition of human vital signs i.e. oxygen level, body temperature, heart rate and blood pressure.
- Wireless transmission of vital signs to the Raspberry Pi using MQTT protocol and database of patient records on Raspberry Pi
- AI based disease prediction on Raspberry Pi and warning messages for the medical doctors
- Chatbot implementation on Raspberry Pi for oral interaction with patients
- Speaker, mic and LCD interface with Raspberry Pi for display of patient vital signs
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 77100 | |||
| Raspberry Pi 4 | Equipment | 1 | 27000 | 27000 |
| ESP32 | Equipment | 1 | 1000 | 1000 |
| MAX30102 (Oxygen and Pulse Rate Sensor) | Equipment | 1 | 1200 | 1200 |
| DS18B20 (Temperature Sensor) | Equipment | 1 | 800 | 800 |
| BMP180 (Blood Pressure Sensor) | Equipment | 1 | 800 | 800 |
| LCD Screen | Equipment | 1 | 12000 | 12000 |
| Hardware Structure | Equipment | 1 | 20000 | 20000 |
| Speaker | Equipment | 2 | 500 | 1000 |
| Power Supply | Equipment | 1 | 3000 | 3000 |
| Mic | Equipment | 1 | 300 | 300 |
| Miscellaneous | Miscellaneous | 1 | 10000 | 10000 |