🔹 Project OverviewThis pilot project aims to explore the use of Edge AI and IoT sensors to improve maintenance efficiency and quality control in a real sugar factory setting in Thailand.We are collaborating directly with a factory that is open to testing an early-stage prototype system. The goal is to use existing PLC data and sensor feeds (such as temperature and pressure in boilers) to develop a lightweight AI assistant that helps: Predict machine conditions Suggest maintenance before failures Visualize real-time and historical sensor data via a dashboard
🔹 Use Case from the Factory (Initial Idea) The factory already uses PLCs to control machines. They want to focus on a critical point in the boiler system. They store some production data but are not yet using it for predictive analysis. They spend millions of THB per year during off-season shutdown for maintenance.Our goal is to help reduce some of this cost using AI-driven insights.
🔹 What We Need NowWe’re looking for someone who can: Design or suggest an edge architecture that connects to PLCs or sensors Work with mock sensor data or propose how to collect/export it Help us develop a small prototype or simulated demo Optionally contribute to the front-end dashboard for visualizations
Your experience in IoT, Edge Computing, and application logic would be a valuable addition to make this pilot real. 🔹 What You’ll GetIf the project is successful and further developed: You’ll be invited as a co-founder or technical lead for the Edge AI track under FabMatch You’ll gain real-world case studies and visibility into Thailand’s industrial applications In the future, we may offer shared revenue, token-based rewards, or even funded contracts You’ll be part of a cross-country team solving practical AI and industry problems
🔹 Collaboration Tools We’ll use Notion for project documents WhatsApp / Telegram for quick discussion Optional GitHub (if we go into prototyping)
🔹 Proposed Timeline (Flexible) | Week | Tasks | |------|-------| | Week 1 | Confirm use case, define data point, finalize sensor strategy | | Week 2 | Set up mock pipeline / simulate data + design AI logic | | Week 3 | Build prototype dashboard or alerts | | Week 4 | Feedback from factory / improve / finalize pilot deck | 🧠 Optional: Questions for YouIf you’d like to move forward, could you help us answer:
Which part do you feel confident helping with? (Data pipeline / AI logic / Dashboard / PLC integration)
Would you prefer working with mock data or helping define the sensor strategy?
Any existing open-source tools or platform you’d recommend?
Are you okay with starting with a “contribution-based” phase and defining rewards later?