SAMMOY Edge

Equipment and control settings for managing connected hardware and operating parameters.
Wash program management for configuring modes, prices, and availability.
Central operator dashboard with wash bays, payments, status indicators, and quick controls.
System configuration page for branding, interface behavior, and operator settings.
Reporting interface with filters and operational performance data.
Detailed operations table for tracking car wash activity and system events.
Tariff and wash program configuration with flexible pricing and time coefficients.
Real-time transaction and payment monitoring dashboard.
Customer kiosk interface with wash programs, balance, timer, and pause controls.

Inspiration

Self-service car wash operators often depend on closed proprietary controllers that are expensive to customize, difficult to maintain remotely, and poorly adapted to local payment systems.

I wanted to build a more open and modular platform that combines equipment control, customer interaction, payments, tariffs, reporting, and remote management in one system.

SAMMOY Edge started as a real industrial automation project for a self-service car wash and evolved into a complete software and hardware platform with the help of OpenAI Codex.

What it does

SAMMOY Edge is an AI-assisted control platform for self-service car washes.

The system provides:

a portrait touchscreen kiosk interface;
configurable wash programs;
balance and remaining-time calculation;
paid pause functionality;
QR payment workflow;
cash payment workflow;
tariff configuration;
operator and reporting pages;
digital input and output integration;
remote administration;
configurable branding for different car wash operators.

The platform is designed to run on an Orange Pi Zero 3 using Linux, Docker, Node-RED, and a browser-based kiosk interface.

How we built it

The system uses a modular architecture:

Orange Pi Zero 3 runs the local application stack.
Node-RED manages control logic, events, device communication, and kiosk workflows.
Docker provides repeatable deployment and service isolation.
Browser-based kiosk UI provides the customer-facing touchscreen interface.
Digital I/O hardware connects the software platform to buttons, relays, sensors, valves, pumps, and other equipment.
Payment workflows support QR and cash-based balance replenishment.
Reporting and configuration pages provide operator visibility and control.

OpenAI Codex was used throughout the engineering workflow. It helped analyze the repository, refactor the interface, debug Node-RED and Docker issues, improve responsive layouts, build payment workflows, review code, prepare documentation, and validate changes.

Challenges we ran into

One of the main challenges was combining industrial control requirements with a modern customer interface.

The system must remain simple for the customer while coordinating multiple internal components, including wash modes, time accounting, payments, pause logic, digital I/O, reporting, and operator settings.

Another challenge was building a responsive kiosk interface for a large portrait display while also supporting desktop and mobile service pages.

We also had to separate hardware-specific logic from reusable application logic so the platform can later support different equipment configurations and customer brands.

Accomplishments that we're proud of

Built a working touchscreen-oriented car wash interface.
Combined payment, tariff, wash-mode, and reporting workflows in one platform.
Migrated the controller architecture to Orange Pi and external digital I/O.
Created a modular system that can be adapted for multiple car wash brands.
Used Codex as an engineering assistant across frontend, backend, deployment, debugging, and documentation tasks.
Prepared a real industrial prototype rather than a purely conceptual demonstration.

What we learned

This project demonstrated that AI-assisted development can be useful beyond conventional web applications.

Codex helped with software architecture, debugging, interface refinement, deployment configuration, and technical documentation while the project still required real industrial automation knowledge, hardware integration, and safety-aware engineering decisions.

We also learned that clear goals, verification steps, and small iterative changes produce better results than large uncontrolled code-generation tasks.