Live status board
Working, complete, waiting, and error refresh together, so anomalies stand out at a glance.
Log in
Machine-reported states stream in from the shop floor: all four states shown live, offline machines detected, alarms raised, and history searchable. Read-only end to end — no control commands are ever sent.
Four machine states
From machine reports and offline detection to alarm tracing — shop-floor signals scattered across the plant, distilled into one real-time view you can count on.
Working, complete, waiting, and error refresh together, so anomalies stand out at a glance.
Machines that stop reporting are flagged automatically and the alarm summary surfaces immediately — no need to watch every machine.
Every state change is recorded end to end and traceable down to the minute, by machine and timeline.
Availability is calculated from real runtime data to expose the bottleneck machines slowing the line down.
Specs
A one-way data path: machine → edge agent → cloud. The platform only reads and only displays — architecturally, it has no ability to send control commands.
A one-way data path. The platform never sends control commands to equipment, keeping monitoring and operation fully separate.
HTTPS, HMAC signatures, and certificate pinning — verified end to end, so data can be proven but not forged.
Pure standard-library Python, compatible with aging shop-floor PCs running Windows 7 and Python 3.8.
Local queueing, automatic retries, and idempotency keys keep data complete and duplicate-free through network jitter.
PostgreSQL stores every state change and event log, queryable by machine, time, and type.
Health checks report service liveness and database version, so collection-pipeline faults are caught immediately.
Get started
An internal company SaaS — sign in for the four-state board, offline alarms, history search, and OEE analytics.