DiaWell: design and development of an artificial intelligence–assisted smartphone application for type 2 diabetes self-management in a resource- constrained South Indian setting

In the bustling cities and quiet towns of India, a silent epidemic is reshaping daily life. Type 2 diabetes, a condition where the body struggles to manage sugar in the blood, affects millions of people across the nation. For those living with it, survival depends on a constant, quiet vigilance: checking blood sugar levels, taking medication on time, watching what they eat, and visiting a doctor regularly. Yet, for many patients in India's smaller cities, this routine is broken by distance, cost, and the sheer busyness of public hospitals. The advanced digital tools that help manage this disease in wealthier nations—sophisticated sensors that track sugar continuously or complex systems that adjust insulin automatically—are often too expensive or require internet connections that simply do not exist in these communities. The result is a gap in care where patients are left to navigate a complex disease alone, often without the timely feedback they need to stay healthy.

This is the challenge that a team of researchers at JSS Medical College in Mysuru, India, set out to solve. They asked a simple but difficult question: how can you build a digital health tool that works for people who have ordinary smartphones, intermittent internet access, and limited budgets? The answer they developed is not a high-tech miracle, but a carefully crafted piece of software called DiaWell. It is a smartphone application designed specifically for the reality of a resource-constrained setting, where the internet might cut out for hours at a time, and where a doctor might see hundreds of patients a day. The researchers did not just imagine this tool; they built it from the ground up, listening to the people who would use it, and they have now shared the blueprint of how it works.

The story of DiaWell begins not in a computer lab, but in the living rooms and clinics of Mysuru. Before writing a single line of code, the researchers sat down with patients, their families, and the doctors who treat them. They asked what was missing. They learned that patients needed a way to log their health data even when they were traveling or in areas with poor signal. They learned that doctors needed a way to see what was happening with their patients between visits, rather than waiting for the next appointment to find out if a treatment was working. They also learned that the language of the app had to be simple, and the interface had to be easy to navigate for people who might not be comfortable with technology. These conversations shaped every decision that followed, turning a list of needs into a concrete plan.

The resulting application is built on a foundation that prioritizes reliability over flashiness. At its core, DiaWell operates on a "offline-first" principle. In many digital health tools, if the internet connection drops, the app stops working or loses the data you just entered. DiaWell is different. When a patient opens the app to record their blood sugar, blood pressure, or weight, that information is saved immediately on their phone, regardless of whether they are connected to the internet. The app holds onto this data securely, waiting for a moment when the connection returns. As soon as the phone finds a signal, the information quietly travels to a central server, where it becomes available to the patient's doctor. This dual-layer approach ensures that the act of managing diabetes is never interrupted by a broken connection, a common frustration in the region.

For the patient, the app acts as a personal health companion. It provides a secure place to log daily readings and set reminders for medication, ensuring that doses are not missed. One of its most distinctive features is an artificial intelligence tool designed to help with diet. Instead of forcing the user to manually calculate the calories in a meal, the patient can simply take a photo of their food. The system analyzes the image and provides an estimate of the calories and nutritional content, offering guidance on portion sizes and food choices. Patients can also upload photos of their prescriptions and laboratory reports, creating a digital record of their medical history that travels with them. The interface is designed to be gentle and clear, offering health tips and showing simple graphs that let patients see their progress over time, turning raw numbers into a story of their own health.

On the other side of the screen is the doctor's dashboard. This is where the true power of the system lies in bridging the gap between visits. Doctors can log in to see a list of all their enrolled patients. They can review the data that has been synced from the patients' phones, spotting trends in blood sugar levels or identifying those who are at high risk of complications. If a patient's numbers are concerning, the doctor can intervene early, sending a digital prescription or a message of advice without the patient needing to wait in a crowded clinic. This continuous loop of information allows for a level of oversight that is rarely possible in routine care, turning a sporadic relationship into a sustained partnership.

The technology behind DiaWell is deliberately chosen to be accessible and reproducible. The team used widely available software tools that are common in the industry, ensuring that the system is not a one-off experiment but a model that others could build upon. The application runs on standard Android phones, the most common type of device in the region, and connects to a secure cloud server. Security is woven into every layer, with the system using a method of digital identification that keeps patient data private while allowing the app to function smoothly even when the internet is spotty. The entire system was developed through an agile process, meaning the team built a basic version, tested it with users, found flaws, and refined it repeatedly until it felt right for the people who would use it.

However, the researchers are careful to define what this paper is and what it is not. This document is a report on the design and construction of the tool itself. It explains how the application was built, how the pieces fit together, and how it functions in a real-world setting. It does not claim that the app has already cured diabetes or proven to lower blood sugar levels in a clinical trial. Those questions are reserved for a separate study that is currently underway. The authors are explicit that the application's ability to improve health outcomes has not yet been measured in this report. Furthermore, they acknowledge that the artificial intelligence used for food estimation is a supportive tool, not a perfect replacement for a professional dietitian, and that its accuracy depends on the quality of the photos and the database it draws from.

The significance of this work lies in its practicality. It demonstrates that advanced digital health solutions do not need to be expensive or dependent on perfect infrastructure to be useful. By focusing on the specific constraints of the environment—intermittent internet, modest devices, and the need for both patient and doctor engagement—the team created a tool that fits the reality of life in a Tier-II Indian city. The architecture they developed, with its ability to store data locally and sync later, offers a blueprint for other developers working in similar contexts around the world. It shows that with the right approach, technology can be adapted to serve the people who need it most, without requiring them to change their world to fit the technology.

DiaWell stands as a testament to the idea that the best solutions are often the ones that understand the ground they stand on. It is a system built not for the ideal world of perfect connectivity and unlimited resources, but for the world as it is. The researchers have provided a detailed account of how they engineered this bridge between patients and doctors, ensuring that the connection remains unbroken even when the signal fades. While the final verdict on its medical impact awaits further study, the creation of DiaWell itself is a significant step forward, proving that effective digital health is possible even in the most resource-constrained settings.