EcoTracker

About

The problem it solves EcoTracker solves the challenge of understanding and reducing individual environmental impact. Many people want to live more sustainably but lack visibility into their carbon footprint across daily activities. EcoTracker provides personalized insights on transportation, energy usage, diet, and shopping habits, offering actionable recommendations to reduce environmental impact while saving money. User Interaction and Data Flow Users interact with EcoTracker through a mobile-first web application with an intuitive dashboard interface. The user journey follows these steps: Onboarding: New users complete a brief lifestyle survey about their transportation, housing, diet, and shopping habits to establish a baseline carbon footprint. Connection Setup: Users can connect various accounts (utility providers, banking apps, transportation services) for automated data collection, or opt for manual tracking. Daily Tracking: The app passively collects data from connected services and allows manual entry of activities through a simple form or scanning receipts. Dashboard View: Users access their personal dashboard showing: Current carbon footprint with breakdown by category Historical trends and progress toward reduction goals Personalized recommendations and potential savings Community comparison with anonymized peer data Challenge Participation: Users can join monthly sustainability challenges to earn rewards and badges while reducing their environmental impact. Data flows securely from user inputs and third-party APIs into our analytics engine, which calculates carbon impact using peer-reviewed environmental models and returns actionable insights to the user interface. The project architecture and development process EcoTracker employs a microservices architecture built on cloud infrastructure: Frontend: React-based progressive web app with responsive design, optimized for both mobile and desktop experiences. Material-UI components provide a clean, intuitive interface with data visualizations powered by D3.js. Backend Services: User Authentication Service: Handles secure login and account management Data Collection Service: Manages API integrations and manual data entry Carbon Calculation Engine: Core algorithm for converting activities to emissions Recommendation Service: ML-powered system generating personalized suggestions Notification Service: Delivers timely updates and reminders Database Architecture: MongoDB for user profiles and preferences TimescaleDB for time-series activity and emissions data Redis for caching frequently accessed data Development Process: Agile methodology with two-week sprints CI/CD pipeline with automated testing through GitHub Actions User feedback incorporated through beta testing program with 50 initial participants Regular performance and security audits The carbon calculation engine incorporates peer-reviewed emissions factors from environmental research institutions, ensuring accurate impact assessments across diverse activities. Product Integrations EcoTracker seamlessly integrates with various services to provide comprehensive tracking: Energy Monitoring: Smart home device APIs (Nest, Ecobee, SmartThings) Utility provider APIs for electricity, gas, and water usage Green Button API for standardized energy data access Transportation Tracking: Google Maps Timeline API for location and travel method detection Ride-sharing services (Uber, Lyft) for trip data Public transit APIs for bus/train usage Strava API for cycling and walking activities Consumption Analysis: Plaid API for anonymized purchase categorization Receipt scanning via OCR and product identification Grocery delivery service APIs (Instacart, Amazon Fresh) Environmental Data: EPA emissions database for accurate carbon calculations Weather APIs to contextualize energy usage Local renewable energy mix data from grid providers Gamification and Social: Apple Health and Google Fit for activity correlation Social media sharing capabilities Community challenge platform Each integration uses OAuth for secure authentication and data permissions, with clear user controls for managing data access. Key differentiators and uniqueness of the project EcoTracker stands out from existing sustainability apps through several unique features: Holistic Tracking: Unlike competitors that focus on single categories (like transportation or food), EcoTracker provides comprehensive coverage across all major aspects of daily life, giving users a complete picture of their environmental impact. Financial Impact Correlation: We uniquely pair environmental metrics with financial savings, demonstrating how sustainable choices benefit both the planet and the user's wallet. Behavioral Science Foundation: Our recommendation engine is built on behavioral science research, using proven techniques like micro-goals, social proof, and habit stacking to drive lasting behavior change. Dynamic Personalization: Machine learning algorithms adapt to individual user patterns, providing increasingly tailored recommendations as more data is collected. Transparent Methodology: Unlike "black box" calculators, we provide clear explanations of how carbon estimates are derived, with links to scientific sources and the ability to adjust calculation parameters. Community Impact Visualization: Users can see the collective impact of the EcoTracker community, reinforcing the value of individual actions when multiplied across thousands of users. Local Context: Recommendations are customized based on geographic location, accounting for regional differences in energy grids, available services, and seasonal factors. Privacy-First Design: Unlike competitors that monetize user data, EcoTracker employs local processing where possible and maintains strict data usage policies. We analyzed similar solutions including Carbon Footprint Ltd, JouleBug, and Good On You, finding that they either lack comprehensive tracking, rely heavily on manual input, or don't provide actionable recommendations tied to financial benefits. Trade-offs and shortcuts while building During our rapid development cycle, we made several strategic trade-offs: API Limitations: For some services without public APIs, we implemented simplified manual entry forms rather than spending development time on reverse engineering or web scraping solutions. Emissions Precision: We initially used averaged emissions factors for some calculations rather than region-specific data. Future versions will implement more granular geographic customization. Mobile Experience: We chose to develop a progressive web app rather than native mobile applications to accelerate time-to-market. Native apps with potentially better performance are on our roadmap. Machine Learning Implementation: The initial recommendation engine uses rule-based logic rather than sophisticated ML algorithms. We've collected training data during the hackathon to develop more advanced models post-launch. Limited Historical Import: Currently, we only support up to 30 days of historical data import from connected services. Expanded historical analysis is planned for future updates. Authentication Simplification: We relied on third-party authentication (Google, Apple) rather than building a comprehensive custom authentication system with features like enhanced 2FA. Testing Coverage: While critical paths have thorough testing, some edge cases have relied on manual testing during the hackathon. Expanding automated test coverage is a post-hackathon priority. Offline Functionality: The current version requires internet connectivity. Offline tracking with background synchronization is planned for a future release. These trade-offs allowed us to deliver a functional MVP within the hackathon timeframe while establishing a clear roadmap for future enhancements. Additional Features Prior to the hackathon, we had developed a basic carbon calculator with manual data entry capabilities. During the hackathon, we added several significant features: Automated Data Collection: Implemented API integrations with utility providers, transportation services, and financial institutions for passive tracking. Machine Learning Recommendations: Developed the initial version of our personalized recommendation engine based on user patterns and regional data. Social Comparison Feature: Added anonymized peer comparisons to provide social context for individual impact. Financial Impact Analysis: Created algorithms to calculate monetary savings associated with carbon reduction activities. Community Challenges: Built the infrastructure for monthly sustainability challenges with progress tracking and rewards. Receipt Scanning: Implemented OCR technology to extract product information from shopping receipts for automatic emissions calculation. Interactive Dashboard: Redesigned the user interface with interactive visualizations showing emissions breakdown and reduction opportunities. Notification System: Added customizable alerts for unusual consumption patterns and achievement celebrations. These additions transformed the project from a simple calculator into a comprehensive sustainability platform with automated tracking and actionable insights.