1. Introduction

Chronic wounds are a major problem for diabetic people around the world. Every year in the United States, about 120,000 amputations of the lower limb not related to trauma are performed on people with diabetes [1]. These amputations happen to people with uncontrolled diabetes because, according to the National Library of Medicine, their high blood sugar causes diabetic neuropathy, where they lose sensation in their nerves. When they get a cut or wound (usually on the foot), they do not notice or feel it, causing the wound to remain untreated and likely infected. This infection, along with the damaged blood vessels, causes the wound to become chronic and not heal, causing the surrounding tissue to die and eventually requiring an amputation of the wounded limb. This commonly happens in the foot; thus, the name of this condition is diabetic foot [2].

Usually, to prevent this, diabetic people are encouraged to go regularly to the doctor to check for these worsening wounds, but for many reasons (usually not enough time or money), they do not. Because of this, many people who have difficulty controlling their diabetes may have a wound that they do not know is worsening and potentially threatening to their limbs.

However, if a wound could be monitored by a doctor remotely, and even could be treated remotely, this could greatly help many diabetic people and prevent them from being disabled from one of their limbs needlessly by giving doctors the ability to give an early diagnosis and medically intervene, curing the wound before it gets worse.

This paper aims to present an IoT-based diabetic foot online monitoring system and early treatment system to solve this problem, as a proof-of-concept design.

2. Background Study

2.1. Scope of Diabetic Foot and Impact of Treatment

Diabetic foot is a condition that causes a wound on a foot to become severely infected and incapable of proper healing, usually requiring the foot to be amputated for the patient’s safety. It is caused mainly by two pathological processes due to diabetes: peripheral arterial disease and peripheral nerve damage. Peripheral arterial disease, in short, causes the blood vessels to thin and narrow causing poor blood supply (especially distal areas like the legs and feet), which can slow or stop the healing of a wound, while peripheral nerve damage causes loss of sensation in nerves, which causes wounds to go unnoticed, allowing for infections of an already unhealing wound to settle in quickly and cause severe damage due to the lack of treatment [3].

Signs of an infection occurring or worsening in a chronic wound are the increase in wound temperature, pus formation (measured by humidity), and in a change of pH, as noted by multiple papers [8], [9], and [10].

2.2. Related Works & Designs

The engineers of FlexiLab at Purdue University used an electrochemical sensor with an Arduino in their design to detect uric acid in the wound and determine its condition [4].

2.3. Electrotherapy Treatment of Diabetic Foot

To treat diabetic wounds, apart from the traditional treatment like liquid antibiotics, there are newer modalities like electrotherapy, which has shown that it can enhance healing of diabetic ulcers/wounds by flowing small amounts of current near the wound site, stimulating immune cells to react and try to heal and fight the infection sooner [5]. These are particularly useful through electronics, which can deliver said small current.

3. Design Specifications

This system is designed to remotely monitor and treat diabetic foot wounds using an IoT-enabled platform. It integrates sensors for real-time measurement of pH, temperature, and humidity, which are key indicators of wound infection. If signs of infection are detected, the system user can actuate electrotherapy or liquid medication remotely via an app-based interface, enabling early medical intervention and reducing the risk of amputation.

3.1. Materials Used

3.1.1. Software Specifications

Firmware Features:

• Sensor data acquisition and calibration

• Wi-Fi transmission to cloud

• Remote control functions (electrotherapy, medication pump)

Table 1. Software Components

3.1.2. Hardware Specifications and Materials

Table 2. Hardware Components

Figure 1. The materials used in this project

4. Methodology

4.1. Creation and Testing of Simulation Model and Monitoring System

To create a model diabetic foot to test and run our design on, we bought a standard fake foot made of silicone. We carved a circular crater (seen in Figure. 2), in which we fitted a plastic dose cup in the crater with modeling clay so our blood simulant (plain water) could be filled in the crater. To represent the electrotherapy treatment, we ran two wires underneath the skin of the foot (also seen in Figure. 2) and attached a blue LED that would turn on when current ran through the wires, representing the current flowing through the wound in electrotherapy. We acknowledge our simulation model is not capable of quantifying or controlling current density and waveform. In future, we plan to incorporate some of these capabilities in our simulation model.

Figure 2. Creation of simulation model. Carving of wound crater (left), and under skin wiring (right)

For our monitoring system, we used a DHT11 humidity sensor placed directly above the wound cup to measure the humidity of wound (caused by pus formation), a DS18B20 temperature probe placed in the wound cup to measure the wound’s temperature, and a PH0-14 E-201-C pH probe placed in the wound cup to measure the wound’s pH, all wired and connected to the Arduino Uno (as shown in Figure. 3 and Figure. 4). The Arduino runs a cloud sketch from the Arduino IDE with the Arduino IoT Cloud Platform, which allows the chip to read values from all sensors and send them, through Wi-Fi, to the Arduino IoT Remote App. The cloud sketch (main program code written by us) is provided as the appendix, and it is available on Arduino website: https://app.arduino.cc/things/495ce5a9-385a-4c64-9eb8-5c8e0aea7259/sketch (requires user login).

Figure 3. Completed assembly of model with Arduino and its temperature, humidity, and pH sensors along with pump and electrotherapy

Figure 4. The breadboard for wiring of sensors and controls to Arduino Uno R4 Wi-Fi

To simulate an infection on our model, we used anhydrous calcium chloride that, when mixed with the plain water in our wound cup, produces a thermogenic reaction that increases the temperature, and humidity (due to water), and changes the pH, mimicking an infection in a worsening wound becoming chronic. These changes are reflected in the app, allowing the user to see the reading of the sensors, which allowed us to evaluate how well our monitoring system was working and what changes were needed.

For the treatment system, the electrotherapy wires and liquid medication (represented by any liquid, we used dyed water) pump are both wired to the Arduino, again shown in Figure. 3, that allows the user to turn each on or off through the app manually. The app also monitors the time that each treatment is on, indicating for what duration each is active, letting us know if each treatment method was connected correctly, through the Arduino, to the app, and told us what part of the treatment system would need rewiring or programming.

4.2. Calibration

We tested our system by checking how our device showed changes in temperature, humidity, and pH values to see if it could properly detect changes that indicated a bacterial infection in a worsening wound.

We did this test by filling the crater with water (a substitute for blood since around 50% of blood is water [7]) and putting all the sensor probes into the crater. The system would check and record the base values on the monitor to show stability.

After adding two teaspoons of calcium chloride to the water due to a reaction with water and calcium chloride producing calcium hydroxide, a strong base, and significant heat, which simulates the bacterial infection, increasing the pH and the temperature and humidity of the wound, we tested the control for electrotherapy to ensure it is working.

As the system finishes recording the data, a check is done to ensure that the value has changed accurately and reflects the change in the pH, temperature, and humidity. If this test had failed, we would need to reiterate the design.

With this, we also learned that our pH E-201-C probe would require a wait time before stabilizing to a correct value, not due to its error range.

The error range of our sensors were as follows:

pH sensor E-201-C error: ± 0.2pH after proper calibration [11], temperature sensor DS18B20 error: the sensors have a quoted accuracy of ± 0.5 Celsius in the range -10 Celsius to +85 Celsius [12], and the humidity sensor DHT11 error: ± 5%RH in the range 20%~90% [13].

4.3. Flowchart Diagram

Figure 5. The flowchart diagram of our system, including both physical hardware and online software

5. Final Result

To use the diabetic foot monitoring system, the pH, temperature, and humidity probes are placed into the simulated wound and are then connected to the Wi-Fi. As the system records the values, they come as analogue numbers on the Cloud server (Figure. 6 and Figure. 7) that are viewed remotely as graphs in the iPhone app, in Figure. 8 and Figure. 9. The humidity was recorded as 38%, the pH of plain water was recorded as 6.9, and the temperature was recorded as 21.19°C. Then we add about two teaspoons of calcium chloride to the water to simulate bacterial infection. The calcium chloride with water forms calcium hydroxide, a strong base, and also produces heat, which simulates the bacterial infection, increasing the pH, temperature, and humidity of the wound. After the addition of calcium chloride, the pH increases to 7.65 and the temperature increases to 32.63°C. The humidity increases to 79%. The changes are noted remotely on the iPhone app, signifying worsening of the wound.

Figure 6. Values of the sensors from the terminal before the addition of calcium chloride

Figure 7. Values of the sensors from the terminal after the addition of calcium chloride

Figure 8. The graph of the values of the temperature (top) and pH (bottom), before and after calcium chloride was added

Figure 9. The graph of the humidity values, before and after calcium chloride was added

Figure 10. Picture of complete simulation model, with monitoring system app on iPhone (bottom-right)

The physician can then promptly initiate early treatment through the controlled electrotherapy and liquid medication pump, for which we have an on-off switch on the iPhone app to control each (please see Figure. 11).

Figure 11. Screenshot of iPhone app with on-off switch for electrotherapy (Blue LED Control), and liquid medicine delivery pump

5.1. Statistical Analysis

To demonstrate reproducibility, enhance validity, and ensure statistical significance, we conducted the test four additional times. The Arduino serial data was compiled into a .csv file and imported into a Python program (supplied as as appendix), which then performed an independent t-test (Welch’s t-test) for each parameter (pH, temperature, and humidity sensor readings). The results for all three sensors in our design are presented in Table 3 and each sensor’s reading are visualized in Figures. 12, 13, and 14.

Table 3. T-Test Results

Figure 12. Graph of pH across the tests

Figure 13. Graph of temperature in Celsius across the tests

Figure 14. Graph of humidity in percentage across the tests

All parameters are statistically significant (all P-values less than 0.05).

6. Conclusions

This study introduces a novel IoT-based simulation platform for the remote monitoring and early intervention of diabetic foot wounds. By incorporating integrated sensing of pH, temperature, and humidity, alongside remotely actuated treatment modules such as electrotherapy and liquid medication delivery, the system is designed to detect key pathological indicators associated with wound infection progression. Experimental testing using a foot model with simulated inflammatory responses demonstrated the system’s ability to accurately record diabetic wound parameters in real time.

The integration of sensor data with cloud-based monitoring presents a practical approach for enhancing physician oversight, particularly for patients with limited access to in-person care. The model’s modular design supports flexibility in future development, and its emphasis on real-time feedback aligns with current trends in personalized and connected healthcare. These results suggest the feasibility of using smart wound management systems to improve outcomes in diabetic care and reduce the incidence of preventable lower-limb amputations.

Future work will focus on proper integration of a more controlled and quantifiable electrotherapy circuit, the miniaturization and enhanced precision of sensors, the addition of imaging capability and artificial intelligence for automated wound assessment, and possible clinical validation. The proposed proof-of-concept model has the potential to contribute meaningfully to the advancement of biomedical engineering solutions for chronic wound management.

ACKNOWLEDGEMENTS

This project received the Gold Medal in the Bayonne District’s Project Innovate competition, and Silver Medal at the 67^th Hudson County STEM Showcase in the year 2025. This project has also been nominated to participate in the Thermo Fisher Scientific Junior Innovators Challenge 2025, to be held in Washington D.C., USA.

The Appendex

MiracleOS_1-7-2025.ino (Main Code)

thingsProperties.h (Library for WiFi)

Python Program used for generating T-Test and Statistical Analysis: