Aetheria

• 

  Title Page

• 

  Overview Page

• 

  Floor Plan Issues

• 

  Floor Plan Analysis w/ AI Suggestions

• 

  Detailed Breakdown of Issues

• 

  ML Powered Graphics

Inspiration

With sustainability at the forefront of global concerns, we recognized that buildings are significant contributors to resource consumption and waste generation. Yet, many facilities lack the granular data needed to make impactful changes. Inspired by the potential of IoT and machine learning to drive environmental efficiency, our team of four set out to create a solution that empowers building managers to make data-driven decisions, reduce waste, and optimize resource usage. We believe that by harnessing technology, we can transform buildings into models of sustainability.

What it does

Aetheria is a Swift-based mobile application designed to revolutionize building sustainability management. While we plan to collect real-time data from strategically placed sensors throughout a building, including trash cans, rooms, electrical panels, water outlets, and HVAC systems, in our current prototype we simulate this data to provide a comprehensive overview of the building's environmental footprint. Key features include:

• Sustainability Dial: Utilizes machine learning linear regression to compute an overall sustainability score, categorizing it as high, medium, or low based on historical and current data trends.
• Priority Ranking: Lists building sections from worst to best in terms of sustainability issues. Users can delve into each section to view detailed problems, AI-powered solutions, estimated offsets, and cost-benefit analyses.
• Interactive Map: Displays precise locations of identified issues within the building, enabling quick action and resource allocation.
• Trend Analysis: Leverages historical data to perform comparative analyses, highlighting improvements or regressions in sustainability metrics over time.

How we built it

We developed Aetheria using Swift for a robust and intuitive iOS experience. Since we haven't implemented the physical sensors yet, we used a mock dataset to simulate sensor readings. All data processing and storage are handled locally within the app, eliminating the need for external backend services at this stage. We implemented the linear regression models directly in Swift, utilizing Apple's Core ML framework for machine learning tasks. The AI-powered suggestions are generated using decision-tree algorithms and predefined rules within the app. We integrated MapKit for the interactive mapping feature to display the locations of identified issues. This approach allows us to focus on developing a seamless user experience and ensures that the app is ready for future integration with real sensors.

Challenges we ran into

One of the main challenges was developing a realistic simulation of sensor data to effectively test our application's functionalities. Crafting accurate and varied data streams required us to understand the expected outputs from different types of sensors and how they would interact over time. Implementing machine learning models without real-world data also posed difficulties; we had to ensure our models are robust enough to handle actual sensor inputs in the future. Integrating multiple components, such as the ML models, backend services, and front-end interface, required careful planning and debugging to ensure seamless operation. Additionally, designing an intuitive user interface that effectively presents complex sustainability data in an accessible manner was a significant challenge that required several iterations.

Accomplishments that we're proud of

We're proud to have developed a fully functional software prototype within the hackathon timeframe. Achieving data visualization and analysis on a mobile platform, even with simulated data, is a testament to our team's dedication and technical skills. The successful implementation of machine learning models to provide actionable sustainability insights, and the creation of an intuitive user interface that makes complex data accessible, are accomplishments we hold in high regard. Our prototype lays a solid foundation for future integration with physical sensors.

What we learned

This project deepened our understanding of mobile app development and machine learning in the context of sustainability applications. We learned how to generate and work with simulated data, which is crucial when actual data is not yet available. The experience honed our skills in integrating different technologies, such as Swift and Python, and reinforced best practices in designing efficient architectures. We gained valuable insights into implementing machine learning models directly within a mobile application using Swift's Core ML framework. We also enhanced our abilities in user experience design, especially in presenting complex data in a user-friendly manner. Collaboratively, we learned the importance of adaptability and effective communication in overcoming project challenges.

What's next for Aetheria

Moving forward, we aim to enhance Aetheria by actually developing and coding IoT sensors to collect real-time sustainability data. Our focus is on implementing these sensors in a real building setting to monitor factors like energy consumption, temperature, and air quality. This hands-on approach will allow us to test the system's functionality, refine our data collection methods, and evaluate the practical aspects of deploying Aetheria in an actual facility. Additionally, we plan to improve our machine learning models with real-world data, enhance the AI-powered suggestions, and possibly expand the platform to support multiple buildings, contributing to broader sustainability efforts.

Built With

• python
• swift
• team