We developed a system that collects live environmental data such as temperature and humidity, sends it to an AI (ChatGPT), and lets the AI decide whether watering is needed. The decision-maker here isn't a fixed set of rules — it's the AI itself. A local web interface connected to the system also allows it to be monitored and controlled remotely. We delivered a working prototype along with the wiring diagrams to our customer.
Our customer's goal was to see whether an experimental system that leaves the watering decision to an AI's flexible judgment, rather than fixed rules, was feasible. We designed and developed the entire process ourselves, from the wiring diagrams and embedded software to the local web interface.
After the needs analysis, we developed the embedded software that reads the environmental sensors, sends the data to the AI, and triggers watering based on its response. We added a local web interface to allow the system to be monitored and managed remotely. We drew up the wiring diagrams, tested the working prototype, and delivered it to our customer.
