bE-More: IoT & Local AI Project Wins Project Day 2025
Discover how bE-More, an energy-saving office system integrating IoT and Local AI, won the Project Day 2025 contest with a secure architecture.
Stock photo for illustration only, not from the actual event
- bE-More won first place at the school's Project Day 2025 contest.
- Combines IoT, Java 23, and Local AI (Mistral:7b) for office energy saving.
- Addresses corporate data privacy concerns by running local AI processing.
- Features a physical office diorama prototype to test real-world sensors and actuators.
The journey of bE-More began in February 2025 as the team prepared for Project Day, the most anticipated annual event at school that transforms the campus into a tech hub where student teams showcase their work to local company representatives. The developers aimed to build a professional solution addressing a real global challenge: energy saving.
To ensure strong market demand, the team looked at the Agenda 2030 guidelines, which challenge companies to meet stricter environmental standards and boost energy savings. However, they identified a major friction point: corporate hesitation to send sensitive telemetries to the cloud for analysis due to strict data privacy concerns.
Stock photo for illustration only, not from the actual event
This challenge led the team to rely on a local AI analysis system using Ollama and the Mistral:7b model, creating a lightweight, fast, and secure solution. This setup allows users to gain data insights and detect anomalies quickly without spending hours reading through complex charts.
The bE-More architecture consists of several communicating components: a Controller built with Java 23 acting as the central desktop hub with WebView dashboards, an IoT Hub powered by ThingsBoard for robust data handling via MQTT on port 8080, Edge Hardware using Arduino to gather real-time sensor telemetries, and an Intelligence middleware (Python & Ollama) bridging ThingsBoard's PostgreSQL database to the AI model.
The integration of Local AI and IoT architectures highlights the rising importance of Edge AI in modern corporate environments. By processing sensitive operational data locally within the private network instead of relying on external cloud servers, organizations can effectively mitigate data breach risks while successfully achieving tangible energy reduction goals in office buildings.
To demonstrate the project during the showcase, the team built a physical diorama simulating an office environment with integrated sensors and actuators. The setup includes workspace LEDs, status feedback lights, an acoustic buzzer, and physical shell buttons to toggle between manual and autonomous modes. Driven by an Arduino UNO R4 microcontroller, the system monitors natural light via a photoresistor and automatically powers down office lighting when light levels exceed 450 while in 'AUTO' mode.
"This double victory proved that our school is a real nursery for innovation, where sustainability and technology meet to solve actual problems."
GiZano
Winning first place alongside fellow classmate project VIPRA highlighted the school's strong innovative environment. For future deployments, the team plans to transition to ESP32 boards for smaller form factors and superior processing power.
Source: Dev.to
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment