About ClimateEdge

Inspiration

Our team was inspired by the challenges faced by farmers in our community. We witnessed how unpredictable weather patterns were affecting crop yields and realized that existing weather forecasting tools weren't precise enough for individual farms. This sparked our idea to create a hyperlocal weather prediction system tailored for agriculture.

What We Learned

Through this project, we've gained invaluable insights into:

The intricacies of IoT sensor networks and data collection
AI and machine learning techniques for weather prediction
The specific needs of farmers and the agricultural sector
Challenges in developing user-friendly mobile applications for rural areas

How We Built It

ClimateEdge was built through a multi-stage process:

Researching and designing low-cost, durable IoT weather sensors
Developing an AI model to analyze and predict weather patterns using TensorFlow
Creating a scalable backend infrastructure with Node.js and MongoDB
Designing an intuitive mobile app interface using React Native
Integrating all components into a cohesive system

Challenges We Faced

Our journey wasn't without obstacles:

Ensuring the accuracy and reliability of our IoT sensors in various environmental conditions
Developing an AI model that could provide accurate hyperlocal predictions with limited initial data
Designing a user interface that's accessible and useful for farmers with varying levels of tech literacy
Balancing the need for frequent data updates with power consumption limitations in rural areas
EMQX Failed us in the data base but we are committed to make sure everything works.

Despite these challenges, we're proud of the solution we've created and excited about its potential to revolutionize agricultural weather forecasting.

Built With

Submitted to

DeKUT Hack

Created by

As the Data Engineer for ClimateEdge, my contributions were crucial in laying the groundwork for our data infrastructure and analytics capabilities. Here's an overview of my key responsibilities and achievements:

1. Data Pipeline Design:
- Architected a scalable data ingestion pipeline to handle real-time data streams from our IoT weather sensors
- Implemented data validation and cleaning processes to ensure data quality and consistency

2. Database Management:
- Set up and optimized our MongoDB database for efficient storage and retrieval of sensor data and weather predictions
- Designed schema to accommodate both time-series data from sensors and structured data for user profiles and farm information

3. ETL Processes:
- Developed Extract, Transform, Load (ETL) processes to prepare raw sensor data for our AI model
- Created automated workflows for data aggregation and feature engineering to support our machine learning pipeline

4. Data API Development:
- Built RESTful APIs using Node.js to facilitate data exchange between our backend, mobile app, and AI model
- Implemented caching mechanisms to improve API performance and reduce database load

5. Performance Optimization:
- Conducted regular performance audits and optimized queries to ensure fast data retrieval, even as our dataset grew
- Implemented data partitioning strategies to maintain system performance at scale

6. Data Security:
- Worked on implementing data encryption and access control measures to protect sensitive farmer and sensor data
- Designed data anonymization processes for any data used in research or shared with partners

7. Troubleshooting:
- Led efforts to diagnose and resolve our backend data fetching issues when we encountered problems with our planned EMQX implementation
- Explored alternative, cost-effective solutions to maintain data flow in the face of budget constraints

8. Documentation and Knowledge Sharing:
- Created comprehensive documentation for our data architecture and processes
- Conducted knowledge sharing sessions with the team to ensure everyone understood our data flow and could work effectively with our systems

Despite the challenges we faced, particularly with our backend infrastructure, my work laid a solid foundation for ClimateEdge's data-driven approach to hyperlocal weather prediction. I'm continuing to explore solutions to our current limitations and am committed to building a robust, scalable data ecosystem for our platform.

Manuel254001 Emmanuel Muchui
Brian Kibiwott
Josphine Mendwa

Updates

Manuel254001 Emmanuel Muchui posted an update — Jul 09, 2024 01:24 PM EDT

ClimateEdge Update: Progress Despite Challenges

Hello ClimateEdge followers! We're excited to share our latest progress:

IoT Sensor Prototype: We've successfully developed a low-cost, weather-resistant prototype for our field sensors. Early tests show promising data collection capabilities.

AI Model Development: Our machine learning model is showing good results in initial tests, accurately predicting short-term weather patterns for specific locations.

Mobile App UI: We've finalized the design for our user-friendly mobile interface. Here's a sneak peek of our main dashboard:

Backend Challenges: We've hit a roadblock with our backend infrastructure. Due to budget constraints, we couldn't implement our planned EMQX-based solution for data management. We're actively exploring alternative, cost-effective options to resolve this. Next Steps:

Secure funding or find an open-source alternative for our backend infrastructure
Begin beta testing of sensors with local farmers
Refine our AI model with real-world data

We're committed to overcoming our current challenges and bringing ClimateEdge to farmers who need it. Stay tuned for more updates!

AgTech #WeatherPrediction #AIforAgriculture #StartupJourney

Log in or sign up for Devpost to join the conversation.

Manuel254001 Emmanuel Muchui started this project — Jul 09, 2024 01:14 PM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.