InspirationEcoPets — Turning Recycling into a World Worth Exploring
Inspiration EcoPets began with a question I had been thinking about for a long time:
If technology can make people walk, explore, compete, and care deeply about virtual characters, could it also inspire them to protect the real world?
Environmental innovation has always been personally important to me. Before EcoPets, I had already worked on an environmental technology concept that explored how artificial intelligence and smart systems could support more sustainable management of natural resources. That experience showed me that technology can help solve environmental problems—but it also taught me that even the most advanced solution depends on people being willing to change their everyday behavior.
Recycling is one of the simplest environmental actions available to almost everyone, yet it often feels repetitive, inconvenient, and disconnected from any visible outcome. People may understand why recycling matters, but they do not always know where to find the correct bin, and a single recycled item rarely feels significant enough to motivate them to continue.
That led to the central idea behind EcoPets:
What if every real-world recycling action helped a digital creature grow?
When I shared this idea with my friend and teammate, Ioanna Serf, it quickly developed into something we both wanted to bring to life. Together, we shaped EcoPets from an early environmental concept into a complete team project combining sustainability, artificial intelligence, geolocation, and game design.
Inspired by the exploration of Pokémon GO and the emotional connection people form with virtual companions, we imagined an experience where the city becomes a playable environmental world—and where progress in the game is earned through genuine positive action.
What It Does EcoPets is a mobile and web application that transforms real-world recycling into an interactive, location-based game.
Its central gameplay loop is:
Discover → Recycle → Scan → Verify → Earn → Evolve
Users can:
Discover nearby recycling points through an interactive map. Identify bins based on the materials they accept. Photograph recyclable objects using their device. Verify materials through AI-powered image recognition. Earn XP and digital rewards for successful recycling actions. Collect, nurture, and evolve unique digital Eco-Pets. Track their progress and environmental activity. Participate in future local challenges, community missions, and leaderboards. Each Eco-Pet is connected to a recycling category. Glass, plastic, and paper can unlock different creatures and contribute to their development. As users continue recycling, their pets gain experience, become stronger, and eventually evolve into rarer forms.
Instead of offering abstract points with no emotional meaning, EcoPets gives every action an immediate and visible result. The user is not simply increasing a number—they are caring for a companion whose growth represents their real-world environmental contribution.
How We Built It We built EcoPets as a team, dividing our attention across the user experience, game mechanics, visual identity, AI scanner, and technical integration while continuously bringing those elements together into one consistent product.
The frontend was developed using React, TypeScript, and Vite, allowing us to create a responsive application that works across desktop and mobile devices. We organized the interface into reusable components and dedicated areas for:
The interactive recycling map AI-powered material scanning Eco-Pet collection and progression XP and reward systems User profiles and activity Mobile-friendly navigation For the backend, we used Python and Flask to process uploaded images, communicate securely with the AI service, and return structured verification results to the application.
Our scanner uses the Gemini Vision API to analyze a photograph and classify the recyclable material as plastic, glass, paper, or an unsupported object. The verification result is then connected to the game logic, allowing a valid real-world action to generate XP and contribute to the progression of the corresponding Eco-Pet.
The map experience was designed around geolocation and recycling-point discovery, creating the foundation for a world in which physical recycling bins become interactive destinations.
The AI is not an additional feature placed on top of the project. It is the mechanism that connects the physical action of recycling with progress in the digital game. Without verification, EcoPets would simply be a game about recycling. With verification, the player’s real-world behavior becomes part of the gameplay itself.
Our MVP demonstrates the most important part of the concept:
The user selects or photographs a recyclable object. The image is sent securely to the backend. The AI analyzes and classifies the material. The application displays a clear verification result. A successful action contributes to the user’s rewards and Eco-Pet progression. Creating Reliable AI Verification Real-world photographs are unpredictable. Lighting conditions, camera angles, complex backgrounds, damaged packaging, and objects made from multiple materials can all affect classification.
We had to convert the AI response into clear, consistent categories that the rest of the application could understand. We also needed to design useful error states for images that were unclear, unsupported, or incorrectly classified. Challenges We Faced Connecting Every Layer of the Application
The scanner depends on communication between the React frontend, Flask backend, AI service, and game state. A problem with an API request, environment variable, dependency, response format, or cross-origin configuration could interrupt the entire flow.
Testing each component individually was not enough. We had to test the complete journey from image selection to verification and reward.
Turning a Large Vision into a Focused MVP EcoPets has the potential to include augmented reality, multiple evolutionary paths, team competitions, school programs, municipal dashboards, sponsored challenges, real-world rewards, and seasonal events.
Our challenge was deciding what to prioritize during a limited hackathon timeframe. We focused on proving the central innovation:
A real recycling action can be verified through AI and transformed into meaningful digital progression. Designing a Fair Reward System
If recycling produces valuable in-game rewards, the system must protect against repeated photographs, unrelated images, and false submissions.
Our current AI classification creates the foundation for verification. Future versions can strengthen it through duplicate-image detection, geolocation validation, timestamps, material-to-bin matching, and verified recycling-point data.
Building as a Team Different parts of EcoPets were developed in parallel, which made communication and version control essential. Integrating the interface, scanner, map, and game mechanics required careful coordination, Git branch management, and repeated testing to ensure that each new feature supported the same product vision.
What We Learned Building EcoPets taught us that environmental impact alone is not enough to create an engaging product. A meaningful application must also be intuitive, trustworthy, emotionally rewarding, and enjoyable to use.
We learned that:
Artificial intelligence works best when it serves a clear purpose within the product. AI responses must be structured, validated, and supported by understandable failure states. Gamification is more meaningful when rewards represent genuine actions. Emotional attachment can turn a simple reward system into a long-term motivation loop. A focused, functional MVP is more powerful than a large collection of unfinished features. Successful teamwork requires communication, flexibility, and a shared understanding of the core vision. Environmental technology can be positive and playful without losing the seriousness of its mission. Technically, we gained valuable experience with React architecture, TypeScript, Flask, API integration, image-upload workflows, AI classification, responsive design, error handling, application state, and collaborative development.
Most importantly, we learned how to transform an ambitious idea into a product that people can see, understand, and interact with.
What We Are Proud Of We are proud that EcoPets uses AI as an essential part of its central experience—not simply as a technical showcase.
The AI creates a bridge between two worlds: the physical world, where a person performs a positive environmental action, and the digital world, where that action creates visible progress.
We are equally proud that EcoPets approaches sustainability through optimism rather than guilt. It invites users to explore, collect, compete, care for a companion, and watch their actions contribute to something larger.
What’s Next Our next steps include:
Improving classification accuracy with more diverse recycling images. Adding duplicate-image and anti-fraud detection. Verifying that each material matches the selected recycling bin. Expanding the Eco-Pet collection and evolution system. Introducing daily missions, streaks, achievements, and leaderboards. Creating school, university, and neighborhood competitions. Partnering with environmentally responsible businesses to offer real-world rewards. Developing anonymized municipal heatmaps and impact dashboards. Exploring augmented-reality Eco-Pet interactions. Supporting smart-bin and IoT integrations. Our long-term vision is to make EcoPets more than a game. We want it to become a shared platform through which citizens, schools, companies, and municipalities can make sustainable behavior part of everyday digital culture.
Pokémon GO used GPS to inspire millions of people to explore the world around them. EcoPets uses GPS, AI, and digital companions to inspire a generation to protect it.
Log in or sign up for Devpost to join the conversation.