- A pre-diabetes management system
For our System Design class, we were tasked with developing a Product-Service-Ecology for a chosen chronic health condition and ideating based on emerging wearable technology into the proposed system.
As I previously helped to design the patch pattern for a laser engraved wearable sensor, I decided to incorporate a non-invasive sweat sensor in my integrated system.
System Mapping, Journey Mapping, User Conceptual Model, Mid-Fidelity Prototyping, Figma, Miro
The project focused on the first three phases of the 4D methodology by defining the current situation of patients, designing the future ideal state of patients, and developing the user interface used by patients.
Diabetes is a disease when your blood glucose level is too high.
Type II diabetes develops when the body becomes resistant to insulin or when the pancreas is unable to produce enough insulin that leads to high blood sugar.
Early-stage Type II Diabetes, also known as Prediabetes, is a high blood sugar condition that is likely to develop into type 2 diabetes. However, the progression can be reversed with lifestyle changes, weight loss, and medication.
To understand the impact of this condition on day-to-day living, I interviewed a senior citizen, Elizabeth, with Prediabetes and mapped out her support system as well as her typical daily routine.
Elizabeth lives with her husband and regularly goes to a community clinic to check her glucose level. Her pet dog Grey was brought home by her son a few years ago.
On a typical day, Elizabeth would be very happy when she walks with Grey, and she would be frustrated when she doesn't know if a food is good to be consumed.
Based on my primary and secondary research, I developed the glucose level management feedback loop for a pre-diabetes patient and identified the key disturbances to the system. The diagram gave me key insights on the features to be considered in my design including understanding food intake and exercise level.
Glucose level is a very abstract concept for prediabetes patients to understand. They need specific information on the impact of data on their everyday life, notably on diet and exercise.
A data-informed mobile application facilitating positive lifestyle change.
The product includes an integrated wearable sensor that monitors glucose level and heart rate, an optimization model of a machine-learning algorithm to predict the change in glucose level, and a mobile app for patients to interact with the integrated system.
Data Optimization Model
The machine learning model explained the key parameters used to establish the data model, the two main target objectives, and how new input data from a new user is helping to optimize the model for personalized prediction.
User Conceptual Model
Having defined the product ecosystem, and the machine learning model, I developed the user conceptual model for the mobile app that both collects input data and provides feedback to the user.
key idea implementation
The following prototyping screens showcased how the integrated system is helping patients to manage their lives through two key questions:
"How will the food intake increase blood glucose level?"
"How could exercise help to bring down glucose level?"
The app allows the patient to see their current Glucose level, log their food and activities, view statistics, and input personal parameters.
Glucose Level Prediction
While logging food, patients will be able to view the glycemic index for different recipes. They can also scan food nutrition information while purchasing.
These food data, combining with the patient's profile data, will be passed into the machine learning prediction model and generate a unique prediction.
Correlation with Lifestyle
In daily statistics, patients will be able to view the relationship between glucose peak and mealtime. They can also see the predicted glucose level without exercise and the differences highlighted.
Through this project, I learned about the feedback loop and also how a mobile app is also part of a larger system. It is important that the system does more than monitor and control, but also it optimizes itself and eventually reaches autonomy. The understanding of the overall system enables me to see how different touchpoints are needed for the user to be engaged. I enjoy this approach to design interactions and I hope to take advantage of this approach for all my future projects.