MachineID sees the Mash Machine 1 pick-and-place cell as one live record — every station and all the data we collect from it in real time. The machine writes a single shared log, Mash sends it exactly as it is, and our Edge agent parses it automatically and maps it onto the live machine record. With no extra work on the Mash side.
A pick-and-place cell with hybrid control. MachineID does not read the PLC directly — it works from one shared log written by the machine control application.
Follow one part from left to right — from infeed through inspection, pick-up, orientation and measurement to sorting. Every station is a real component from the machine log, and each one shows which data we read from it.
Four vibrating plates shake the mixed parts apart and present them flat to the cameras.
A camera above each feeder finds the right part and checks it before the arm picks it up.
The gantry arm takes the part from the feeder while the measuring jaws check its dimension against tolerance.
A rotary table turns the part into the correct position and it is then picked up again.
Good parts are placed into 8 positions per tray (2×4). Trays A and B fill and empty alternately.
A part of the wrong type or with a defect is automatically rejected to waste.
The machine writes one shared log. Mash sends all of it, exactly as it is. Our Edge agent parses it, translates it and maps it onto the machine record — with no work on the Mash side.
The .NET application of the machine records every movement, pick, measurement and fault into one shared log.
It sends the log as it is over the Mash VPN — no filtering, no edits, no export format.
It reads the tags in the log (including the Slovak terms), filters out the noise, keeps only real alarms and maps them onto the MachineID model.
It receives the translated signals and routes them into the correct machine record.
A live view of the machine — run state, parts, quality, trays, faults, counters.
Nine live signals from the machine are grouped into four value themes — and each one answers a specific operational question.
Live run state and running time, every good part counted the moment it is made and tagged with the product type, plus current throughput per shift.
Every part is dimensionally checked in the measuring gripper and captured by four cameras — so you see the scrap rate, which product is failing and the trend of the measured dimension.
Real faults are reported the moment they occur. The PLC connection state is monitored too — dropouts tend to precede a stop, so you get an early warning, not just an alarm after the fact.
Cycle time per tray shows where the cell slows down; pick positions and insertion failure counts pinpoint the feeder or position causing trouble — a rising failure count is an early sign of wear.
Twelve real examples answered straight from the Mash Machine 1 — with no dashboards to learn.
Mash sends the whole log exactly as it is — and MachineID turns it into a live view of the machine: production, quality, faults and maintenance. That same live data is the basis for live dashboards, fleet-wide views and prediction before something fails. Adding another machine is a small configuration entry, not a new project.
