Manufacturing Operations Platform
Manufacturing · Automotive & Aerospace Suppliers
2024
enterprise
What they needed
The business runs five plants producing precision metal components for automotive and aerospace suppliers. Each plant tracked production on a mix of paper travelers, spreadsheets, and a 15-year-old MES module that only synced overnight. Line supervisors had no way to see machine status or order progress without walking to a terminal, so downtime often went unreported for hours. Corporate could not see cross-plant capacity in real time, which meant late orders were caught only after a customer called. Automotive and aerospace customers also expected traceable, audit-ready downtime and root-cause records on demand, which the paper-based process couldn't produce quickly. A new ERP rollout was six months away and couldn't fix the floor-level gap on its own.
What we built
Prixelo built a production-tracking platform that ingests machine and line data — captured through retrofitted IoT sensors on the existing equipment — through Kafka event streams and stores it in PostgreSQL for reporting and audit history. A React dashboard gives plant managers and corporate leadership a live view of throughput, changeover time, and downtime causes across all five plants, with drill-down to a single work order. Alongside it, we built a mobile companion app for line supervisors: a React-based interface they carry on the floor showing real-time machine status, open alerts, and shift targets, with push notifications when a line stops. The Node.js backend runs as a set of Docker services per plant, so a network outage at one site never takes down the others. Supervisors log downtime reasons directly from the app instead of a paper traveler.
How we approached it
Chose Kafka for machine event ingestion — fed by retrofitted IoT sensors on existing equipment — so downtime and cycle-time events stream in real time instead of the nightly batch syncs the old MES relied on.
Deployed the Node.js backend as per-plant Docker containers behind a shared gateway, so one plant's network outage never took down monitoring at the other four.
Built the supervisor app as an installable React PWA rather than two native codebases, so IT could push updates to floor tablets without app-store review delays.
Modeled downtime reason codes and work orders in PostgreSQL with an audit-friendly schema, so root-cause reports could be pulled directly for customer quality audits.
The impact
Unplanned Downtime Reduced
Downtime Reporting Speed
Machines & Lines Connected
"We used to hear about a line stoppage the next morning, during the overnight MES sync. Now a supervisor flags it from the app on the floor and corporate sees it in the same minute. That change alone paid for the project."
VP of Operations
Manufacturing company
Tell us what
to build.
Bring us the one everyone said was too hard. Free, no-obligation reply within 24 hours — NDA on request.