R&W Tech: the conductor of fifty welding machines
Fifty laser welding machines that never spoke to each other, now coordinated by a single central brain.
- Year
- 2024
- Client
- R&W Tech srl
The problem
Fifty precision laser welding machines, each excellent at its own job and completely unaware of what the others were doing. No coordination, data locked inside the individual microcontroller, maintenance by guesswork and performance analysis effectively non-existent.
R&W Tech asked us for a central brain: something that would watch every machine, get them communicating and coordinating, without losing the granularity of the data. Not the average across the line: every single weld point.
How we solved it
The system is a microservice architecture built for industrial reliability: an API layer for the business logic, a daemon that coordinates the process, a dedicated worker for each welding machine, and an event bus that keeps the components talking without coupling them together.

Adding a machine to the line means entering an IP address: the daemon recognises the new node and spawns its dedicated worker by itself. Automatic discovery is the part of the project we are happiest with, precisely because you never see it.
Everything ships as a ready-made virtual machine on VMWare, VirtualBox and Hyper-V: backups by snapshot, over-the-air updates that arrive without stopping production, and a licensing system that governs access.
The results
Today R&W Tech no longer manages fifty isolated machines: it conducts a production ecosystem that watches itself work and corrects itself accordingly.
At the heart of the system is a microservice architecture built for industrial reliability: React for the real-time dashboard, Node.js and TypeScript for the API, workers and daemons in an event-driven style, Go for the OTA update manager and licence handling, MongoDB for business logic and welder data, Redis as cache and event bus, NGINX as reverse proxy, all containerised with Docker on a Debian base. Communication with the microcontrollers goes over REST APIs, dashboard updates over WebSocket, with ingestion up to 3.33 Hz per welder. Shipped as a virtual machine (VMWare, VirtualBox, Hyper-V) with everything preconfigured: no setup, backups by snapshot.
Stack: React, Node.js, TypeScript, Go, MongoDB, Redis, NGINX, Docker
At the heart of the system is a microservice architecture built for industrial reliability: React for the real-time dashboard, Node.js and TypeScript for the API, workers and daemons in an event-driven style, Go for the OTA update manager and licence handling, MongoDB for business logic and welder data, Redis as cache and event bus, NGINX as reverse proxy, all containerised with Docker on a Debian base. Communication with the microcontrollers goes over REST APIs, dashboard updates over WebSocket, with ingestion up to 3.33 Hz per welder. Shipped as a virtual machine (VMWare, VirtualBox, Hyper-V) with everything preconfigured: no setup, backups by snapshot.
Stack: React, Node.js, TypeScript, Go, MongoDB, Redis, NGINX, Docker
- Year
- 2024
- Client
- R&W Tech srl
Got something similar in mind?
Tell us what you want to build and we will help you find the shortest way there.







