Still Loading...：以 Arduino UNO Q 打造的實體 AI 任務伴侶

⏳ Still Loading... — Your Downloads, Without the Constant Checking.

Ever started a download, build, backup, or other long-running task and found yourself repeatedly checking whether it's finished?

Meet Still Loading..., a physical AI-powered task companion built with the Arduino UNO Q.

💡 The idea: Instead of constantly staring at a progress bar or refreshing a terminal, let a small desk-side device keep an eye on your task and tell you when something deserves your attention.

🔧 How it works

• 📊 Displays task progress on the UNO Q's LED matrix.
• 💓 Receives heartbeat and progress updates from a Mac-side agent.
• 🟠 Detects slow progress and suspected stalls using a state engine.
• 🧠 We're integrating a compact Gemma model running locally on the UNO Q to interpret recent progress and speed patterns and provide concise explanations.
• 🟢 Uses physical LED feedback to distinguish running, slow, stalled, failed, and completed states.

🚀 Why it matters

The goal is simple: make waiting more effortless. Your task keeps running, while Still Loading... handles the monitoring and gives you a physical signal when you need to pay attention.

Rather than relying on cloud AI, we're exploring on-device inference so the device can interpret task behavior locally.

Built around the Arduino UNO Q, combining Linux-side processing with physical hardware feedback.

The project is still evolving, and we're working toward making local AI reasoning and real download monitoring part of the complete experience.

Arduino #ArduinoUNOQ #Gemma #EdgeAI #AI #EmbeddedSystems #StillLoading #HardwareHackathon