Plantagotchi

Plant with hardware
Arduino graph
Healthy Plant
Thirsty Plant
Dizzy Plant

Inspiration

We grow up learning how to care for pets or humans around us but rarely about how to care for the life quietly thriving around us. Plantogatchi reimagines that connection by turning plant care into an interactive experience , one where nature finally speaks back. It’s not just about keeping a plant alive; it’s about learning to listen, bond, and grow alongside it.

What it does

Plantogatchi gives your plant a voice. Using bioelectrical sensors, it listens to the tiny signals plants naturally produce. Changes caused by water levels, humidity, temperature, and gas levels and translates them into emotions and visuals you can understand.

It turns plant care into an interactive experience: your plant reacts when you water it, temperature changes affect its mood, and you can track its overall “well-being” through an intuitive app or display.

How We Built It

We built Plantogatchi around an Arduino UNO R3 microcontroller connected to multiple environmental and bio-signal sensors:

MQ-3 – Detects air quality and gas concentration
HL-83 – Detects presence of liquid water
YL-69 – Monitors soil moisture levels
DHT11 – Captures ambient temperature and humidity
BioAmp EXG Pill – Senses subtle wind-induced bioelectrical variations from the plant

All sensors interface through the Arduino and stream real-time data to a Processing-based backend via WebSockets.
There, data cleaning and optimization algorithms process incoming signals for stability and interpretability.

The processed results are then transmitted to our React Native mobile app,
where we visualize the plant’s emotions and responses using Krita-based hand-drawn animations.

This full pipeline from sensors to backend to animated interface allows Plantogatchi to turn raw plant signals into an expressive, interactive digital experience.

Challenges we ran into

One of the biggest challenges we faced was dealing with signal noise from the BioAmp EXG Pill.
The raw bioelectrical data was extremely noisy, so we had to perform multiple calibration tests to determine reliable value ranges for consistent readings.

Another major challenge was finding suitable APIs for plant care requirements. Most publicly available APIs were either paid, incomplete, or designed for unrelated use cases, which made integrating real-time plant insights into our system more difficult than expected.

Accomplishments that we're proud of

Designed visually appealing, hand-drawn animations that are charming, expressive, and engaging for all age groups.
Achieved smooth integration between live sensor readings and animations, ensuring the visuals react consistently with real-time data.
Delivered a convincing and well-structured business pitch, clearly communicating the vision and potential of Plantogatchi as an interactive, educational, and eco-conscious platform.

What we learned

How to make multiple sensors work in coherence, ensuring stable and synchronized data flow across different environmental and bioelectrical inputs.
The process of ntegrating real-time sensor data with interactive animations in a web or mobile app to create a seamless, responsive experience.
That building a reliable and sellable business pitch takes much more than just a good idea, it requires clear communication, strategy, and market understanding.

What's next for Plantagotchi

Enhanced UI/UX: Expanding the app with more interactive animations, additional plant types, and virtual companions to boost user engagement.
Gamified Experience: Introducing a point-based streak system inspired by Snapchat to reward consistent plant care and learning progress.
Improved Hardware: Developing a more compact, sustainable, and higher-precision sensor module capable of capturing richer environmental and bioelectrical data.
Educational Integration: Adapting Plantogatchi for classroom use, making plant care and bio-interaction an accessible, hands-on learning experience for students of all ages.

Built With

arduino
bioamp
dht-11
hl-83
krita
ldr
mq3
pl@nt.net
react-native
typescript

Submitted to

natHacks 2025
- Winner 3rd Place - Standard Division (Multiple Tracks)

Created by

I led the AI/ML integration and served as the project manager and presentation lead for Plantogotchi. I oversaw the entire workflow from ideation to final submission, ensuring smooth coordination between the hardware, software, and design teams. My role focused on organizing tasks, aligning goals, and keeping the team on track throughout the 64-hour hackathon.

I designed and delivered the final pitch presentation, created and refined all slide decks, and developed our business model that connected both the technical and social impact aspects of the project. I also managed the Devpost submission, carefully adjusting and editing each section to make sure our story, visuals, and technical explanations were cohesive and polished.

Alongside management and presentation, I contributed ideas for integrating AI and ML models into our system, exploring how sensor data could be used to predict environmental stress and enhance plant care feedback.

Raahim Khan
I was responsible for deep research, solution framing, and presenting our project to the judges. I conducted extensive research into plant physiology, existing smart-plant products, and relevant AI/ML approaches to help shape our core problem statement and validate the feasibility and impact of our solution. This research directly informed how we positioned Plantogotchi as both a technically innovative system and a meaningful, user-centric product.

I designed and built the pitch deck from the ground up, translating complex technical concepts—such as bioelectrical plant signals, sensor-driven insights, and AI-powered interpretation—into a clear, compelling narrative that judges could quickly understand. I focused on storytelling, visual clarity, and logical flow to highlight our motivation, technical architecture, social impact, and future potential.

During judging, I served as the primary presenter, explaining our solution end-to-end, answering technical and conceptual questions, and clearly articulating our value proposition. My role ensured that the depth of our work, the novelty of our approach, and the cohesion of the team’s efforts were communicated effectively and confidently under pressure.

Reza Abdi
I led the project development. On hardware side I worked on integration and data pipeline for our project. I ensured flawless communication between the Arduino and multiple sensors, including the BioAmp EXG Pill, MQ-3 alcohol sensor, DHT11 temperature & humidity sensor, and raindrop sensor. I also conducted experiments to discover optimal environmental operating conditions for plants, helping establish reliable baselines for sensor calibration and system performance.

On the software side, I programmed the Arduino to collect and transmit sensor readings in real-time to our app backend via WebSocket, ensuring seamless live data flow. I also researched existing machine learning models and plant-related APIs to benchmark and guide our model development.

Throughout development, I contributed rigorously to app feature testing and QA, diagnosing integration issues early and optimizing user experience. As the main feature advisor, I provided continuous input on functionality, sensor-driven insights, and app hardware synergy to keep our product vision consistent and technically feasible.

Muhammad Abdur Rab Siddiqui
I led the animations and illustrations for Plantagotchi, focusing on visualizing real-world factors like temperature, soil moisture, humidity, watering, and wind. I created responsive animations — leaves swaying, pots shifting, plants reacting — to make the app feel alive and emotionally engaging while keeping a clean, consistent style.

I also helped integrate the PlantNet ML API to identify plant species and personalize care needs in real time. This AI-driven system dynamically adjusted visuals and feedback — for example, a cactus and fern would react differently to overwatering — blending intelligent recognition with immersive, adaptive storytelling.

Muhammad Zain Asad
As the lead full-stack developer for Plantogatchi, I was responsible for connecting every layer of the system—from hardware sensors to the interactive app interface—to create a seamless, emotionally responsive experience. I built and maintained the real-time communication bridge between the Arduino microcontroller and our Processing-based backend using WebSockets, ensuring stable and continuous streaming of data from the BioAmp EXG Pill, DHT11, MQ-3, HL-83, and YL-69 sensors. I implemented signal calibration and filtering routines to reduce noise and improve the interpretability of the bioelectrical readings, which were crucial for reliably expressing the plant’s “emotions.”

On the data side, I oversaw the preparation and validation of the datasets used in our ML models to classify plant mood states, ensuring that the data fetched from external APIs and live sensors aligned with the expected ranges and formats. I designed the backend to process and clean incoming data before forwarding it to the app, guaranteeing low-latency updates and consistent emotional feedback.

On the front end, I worked on integrating the Krita-based hand-drawn animations into our React Native app, carefully tuning the timing and transition logic so that the visuals reacted instantly to sensor changes without lag or jitter. I also ensured the app communicated efficiently with the backend WebSocket server, managing state synchronization, reconnection protocols, and user experience flow to maintain smooth performance across devices.

Overall, my focus was on making the technical pipeline—from raw bioelectrical input to expressive animated output—feel natural, responsive, and alive. Through the hackathon, managed to orchestrate multiple hardware and software systems into one cohesive experience, balancing engineering precision with creative design to make plants truly “speak” in a way humans can feel and understand.

Ahmed Zafar
I shaped the frontend and experience architecture for Plantagotchi, focusing on how raw sensor data and AI outputs were translated into an emotionally engaging product.

I hand-illustrated plant animations and defined how plant states, transitions, and micro-interactions should respond to environmental inputs such as moisture, temperature, and care events. The goal was to create visuals that felt expressive and intuitive rather than purely decorative.

On the frontend, I established animation state logic, timing, and transition patterns so visual feedback updated smoothly and consistently as live data changed. This included making decisions on how different plant species should visually communicate stress and recovery while maintaining a cohesive style across the app.

I also contributed to shaping how plant identification influenced frontend behavior, ensuring that once a species was recognized, its visual responses aligned with real-world characteristics. My work bridged frontend engineering, design direction, and user experience to make complex system behavior immediately understandable to users.

Ibrahim Mansoor
Muhammad Salaar Butt