Jonathon Lee Wei Jie

Responsible, diligent, and adaptable individual with a strong willingness to take on new challenges and responsibilities. Currently pursuing a Bachelor’s degree in Engineering Systems and Design (Class of 2027) at the Singapore University of Technology and Design (SUTD), with a planned specialization in Business Analytics and Operations Research (BAOR).

Fluent in English and Chinese, with basic proficiency in Bahasa Indonesia. Awarded the SUTD Undergraduate Merit Scholarship in recognition of academic excellence and leadership potential.

Proficient in R programming, Python, SQL, Excel and Power BI, with growing interest and experience in coding, optimization and data analytics. Eager to apply analytical and technical skills to solve real-world problems, while continuously exploring opportunities to develop professionally and personally.

1. Risk Prediction Model for Escalation of Care – CGH@Home (Changi General Hospital)
Collaborated on a data analytics project to develop a predictive model that identifies patients at risk of clinical deterioration (readmission) within the CGH@Home program, which delivers hospital-level care to patients at home. The project involved comprehensive data cleaning, exploratory data analysis, and feature engineering using Excel, Power BI, R and Python.

Faced with challenges such as a small, imbalanced dataset and missing values, we employed median imputation and categorical placeholders to ensure data consistency. To enhance model performance, we generated synthetic data and applied SMOTE (Synthetic Minority Over-sampling Technique) to address class imbalance.

Utilized machine learning techniques including Random Forest and CatBoost classifiers to train multiple models. The final model, trained using selected features and tuned parameters, achieved strong performance with an AUC-ROC of 0.97, F1 Score of 0.90, and precision of 98% (non-readmission) and 84% (readmission).

This model acts as a clinical decision-support tool, enabling early intervention and helping reduce unplanned readmissions, ultimately improving patient outcomes and optimizing hospital resource allocation.

2. Hospital Meal Optimization – Integer Linear Programming Project (SUTD)
Formulated and solved an Integer Linear Program (ILP) to minimize hospital food costs at Changi General Hospital while meeting nutritional requirements for 1,071 inpatients. Modeled a simplified meal plan using rice, chicken drumsticks, and eggs to satisfy daily macronutrient needs (carbohydrates, protein, fats) under realistic dietary and operational constraints.

Implemented the model using Julia, JuMP, and MOSEK solver. The optimal solution provided exact food quantities and achieved a minimum daily cost of SGD 6,827.95 while ensuring balanced and realistic meal composition. Incorporated binary and Big-M constraints to enhance solution feasibility and practicality

3. Lightscape Pavilion @ Changi City Point – DTI (Design, Thinking & Innovation) Project
Led and co-developed a multi-phase urban design project aimed at enhancing social interaction and spatial usability through responsive lighting and sustainable architecture. The project began with individual field research, where I analyzed underutilized urban spaces, created site models, and proposed improvements. I later proposed a concept for the area outside Changi City Point, featuring interactive mini domes with karaoke, TV pods, and light-based encouragement messages, supported by a social café space to foster community engagement.

Building on this, our team collaboratively designed the Lightscape Pavilion, a parametric shelter structure inspired by leaf stomata. Using biomimicry and natural patterns, the pavilion integrates smart glass panels equipped with motion, photovoltaic, and light sensors. These enable dynamic lighting effects and solar energy usage, creating an ambient environment that shifts throughout the day. By balancing private and communal zones, the pavilion addresses the lack of seating, gathering spaces, and visual interest in the area, encouraging both spontaneous and intentional social interactions.

Key features include:

• Interactive lighting system that responds to movement and sunlight.
• Smart glass modules for light modulation, enhancing comfort and engagement.
• Sustainable design approach using solar-powered components and modular architecture.
• Public space revitalization aimed at inclusivity, emotional uplift, and community bonding.

This project demonstrates an integrated approach to urban design, technology, and human-centered thinking to transform an overlooked public space into a dynamic and socially engaging environment.

4. Solar-Powered Food Dehydrator – DTP III Project (SUTD)
Co-developed a low-cost, solar-powered food dehydrator designed to address food waste and promote sustainability in Singapore. Conducted research on national food waste issues and environmental data (solar irradiance, humidity, temperature) to evaluate the feasibility of solar dehydration.

Built a functional prototype using a 20W solar panel, dual DC fans, and a lightweight frame with acrylic and reflective materials to maximize heat retention. Performed hypothesis testing using five-year solar intensity data and experiments measuring water removal efficiency under real conditions. Achieved average dehydration effectiveness of 3.33% (4.14% without outliers), with a confirmed operability rate of 99% confidence using Z-test statistical analysis.

Performed cost analysis showing a prototype build cost of ~SGD 175.65, with scalability for mass production. The design emphasizes energy efficiency, food hygiene, and portability, aligning with Singapore’s solar energy targets and circular economy goals

5. Solar-Powered TV Remote – SUTD Design Energy Systems Project
Designed and prototyped a sustainable alternative to conventional battery-powered TV remotes by integrating a rechargeable Li-ion battery and compact solar panel. Conducted battery capacity and solar panel sizing using real power consumption data, energy calculations, and peak sun hour estimates in Singapore.

Selected a 3.7V 3400mAh Li-ion battery and a 5V 400mA solar panel to ensure sufficient autonomy and charging capability. Validated performance through multimeter testing, energy analysis, and solar simulation under 416 W/m² lighting. Despite increased size (~869% volume increase), the modified device successfully maintained remote functionality while reducing reliance on disposable batteries and minimizing e-waste.