Industry 4.0 on a carton erecting machine is not about buzzwords. It is about three measurable things: knowing exactly why the line stopped, when parts will wear out, and having the machine tell you instead of waiting for a jam. ZRAY builds the carton erecting machine with IoT sensors on the same platform that runs Mars and Duracell packaging lines, with 66 patents behind the mechanics and an IIoT layer that we ship, commission, and support.

This article covers what sensors are actually installed, what data the machine generates, how it reports to your SCADA/MES or cloud, what the OEE dashboard looks like in practice, and how the same IoT connection delivers remote diagnostics and firmware updates. We keep it concrete — sensor lists, data formats, alarm handling — because that is what an engineer needs to size and justify the investment.

What Sensors Are Installed on the Machine

The IIoT layer monitors the same systems a technician would check manually, but continuously. Standard sensor set on every ZRAY machine:

The machine controller (Siemens PLC) reads all of these at cycle rate. The IoT edge module aggregates the data and decides what stays local (alarms, safety) and what goes up to the network.

Data Generated: What You Can Actually See

The machine generates three data layers:

1. Alarm log — every alarm code with timestamp, shift, and cycle count. This already exists on the HMI; the IoT layer syncs it to the network so it survives operator resets.
2. Performance counters — total cycles, successful cycles, jam cycles, average cycle time per shift. From these, OEE components are derived: availability (uptime vs. scheduled), performance (actual vs. rated CPM), quality (boxes formed within tolerance).
3. Baseline and drift tracking — the machine stores a commissioning baseline (cycle time, vacuum range, glue temperature curve). When a value drifts beyond a threshold, the machine flags a predictive alert before the failure becomes a jam.

This is the difference between “the machine jammed at 14:22” and “the vacuum pressure has been dropping 0.5 kPa per day since Tuesday, the cup set is due for replacement at 6,000 hours.” That second statement is the one that saves money.

Reporting to Your System

Integration options, in increasing depth:

• OPC UA — the standard IIoT interface. SCADA or MES connects to the machine’s OPC UA server and reads the tag structure: cycle count, alarms, status, parameters. This is the option we recommend for plants with an existing SCADA layer.
• MQTT — the machine publishes JSON payloads to your broker for cloud dashboards. Tag structure is documented in the machine manual; sample payload below.
• Modbus TCP — for legacy line controls that cannot run OPC UA. Read-only access to the core registers.
• ZRAY cloud portal — if you do not want to run your own stack, the machine connects to our portal over a secured VPN, and you get the OEE dashboard and remote diagnostics with no local server.

Example MQTT payload (JSON):

{
  "machine": "ZRAY-SPE-40",
  "ts": "2026-08-24T09:15:02Z",
  "cycle": 184322,
  "status": "RUNNING",
  "alarm": "none",
  "vacuum_kpa": 38.2,
  "air_mpa": 0.62,
  "glue_degc": 162,
  "avg_cycle_s": 3.1,
  "oee_shift": 0.86
}

Every field is documented in the manual. The tag names do not change between firmware versions, so your dashboard does not break on updates.

OEE Dashboard: What to Track First

Do not build a 20-tile dashboard on day one. Start with five tiles that map to real line problems:

• OEE per shift — the headline number. Target 85%+ on a one-shift erector line; the gap is usually a mix of changeover, jam, and waiting-for-blanks time.
• Jam count per shift — tracks the effectiveness of your feeding system maintenance. Rising jam count is your earliest warning of cup wear.
• Alarm code histogram — which code dominates this week. If AIR-LOW climbs, the compressor or filter is the problem, not the machine.
• Changeover time — logged per size change, trended weekly. Creeping changeover time means worn cranks, loose guides, or operators skipping setup steps.
• Vacuum and glue trends — the predictive pair. A falling vacuum trend and a glue-temperature curve that overshoots are both early failure signals.

The same dashboard works from the ZRAY cloud portal if you take that option, with role-based access for operators, maintenance, and management.

Remote Diagnostics and Support Loop

The IoT connection is not just for your engineers. It closes our support loop:

• Remote access: with your permission, our engineers connect over a secured tunnel, read the alarm history, pull the parameter baseline, and compare it against commissioning values. Most issues resolve without a site visit.
• Firmware updates: we deploy parameter and firmware updates remotely, with your approval, outside production windows.
• FAT and commissioning data: your machine’s acceptance data is stored in its profile. When you open a support case, we pull the profile instead of asking “what were the settings at handover?”

The security model is explicit: the machine initiates outbound connections only; no inbound ports are opened on your network; all traffic is encrypted; access requires your authorization per session. We do not store production recipe data on our servers.

Parameters Reference for the IIoT-Equipped Machine

Implementation Steps for Your Plant

Get from purchase to a live OEE dashboard in five steps:

1. Specify integration at quotation: tell us your SCADA type or broker, and we wire the interface at build. Retrofitting is possible but costs more than doing it at the factory.
2. Commission the data path: at installation, we verify the OPC UA/MQTT connection, tag map, and dashboard tiles against your IT contact.
3. Baseline week: run one week, capture baseline cycle time, vacuum and glue curves. This becomes the drift reference.
4. Set alert thresholds: thresholds are set from the baseline week, not from theory — a vacuum threshold that is right for one plant is wrong for another.
5. Train the shift team: operators learn to read the five tiles and respond to predictive alerts, not just to alarms.

Security and IT Handover Checklist

The IoT layer works only if your IT team signs off on it. Here is the handover checklist we walk through with your IT contact at installation:

• Network placement: the machine sits on the production network, not the office network. It initiates outbound connections to your broker (MQTT) or answers OPC UA requests from your SCADA — no inbound ports opened on your firewall.
• Certificates and credentials: machine certificates and broker credentials are provisioned at commissioning, not shipped with a default password. We document the certificate renewal procedure in the handover pack.
• Session-based access: remote diagnostics sessions require your explicit authorization per session, with a time window and a session log you can review.
• Data retention: the machine keeps 12 months of local data; what your SCADA stores is your decision. We do not store production recipe or order data on our servers.
• Firmware change control: updates are deployed outside production windows, with your approval, and the machine rolls back to the previous version if a post-update self-test fails.

Two practical notes from plants that run the ZRAY cloud portal: assign one IT owner for the machine’s network credentials, and schedule a quarterly connectivity check together with your maintenance calendar. A machine that loses its data link silently runs fine mechanically, but you lose the predictive alerts — which is the whole point of the IIoT layer.

Why This Pays Back

Concrete numbers from our installations: plants that act on predictive alerts (vacuum trend, glue overshoot) reduce erector-related downtime by 15–30% in the first season, and shift OEE on the erector line from the mid-70s to the mid-80s. The sensor set adds a small percentage to machine cost; the downtime it removes pays for itself inside the first harvest or production cycle.

For the full IIoT specification on a high-speed platform, see the Double-Head High-Speed Case Former with servo axes that report position and torque data. On a single-lane line, the Single-Piece Case Erector carries the same sensor set and connectivity.

Tell us your SCADA/MES environment and target CPM. We will confirm the interface, the tag list, and the dashboard tiles — email [email protected] or WhatsApp +86 13681839278, with a reply within 24 working hours. Start the technical discussion on our contact page.