# IoT Shop Floor Integration: Connecting Machines to Your Manufacturing ERP
The gap between what happens on the shop floor and what your ERP displays is the root cause of most manufacturing inefficiencies, as Gartner's manufacturing technology research has consistently highlighted. Operators know a machine has been down for an hour, but the ERP still shows the order as on schedule. IoT shop floor integration closes this gap permanently.
The Visibility Gap in Manufacturing
Traditional shop floor reporting relies on manual data entry — shift reports submitted hours after events occur, production counts rounded or estimated, and downtime reasons recorded from memory. This creates a dangerous information lag:
What Manual Reporting Looks Like
- Shift reports entered at end of shift — 8 hours of latency minimum
- Production counts estimated from batch totals — accuracy of 85-95% at best
- Downtime events logged retroactively — root causes lost or misremembered
- Quality data transcribed from paper forms — transcription errors are common
What IoT-Integrated Reporting Delivers
- Real-time machine states updated every second — no information lag
- Automatic part counting from machine signals — 99.5%+ accuracy
- Instant downtime capture with fault codes — every event logged with context
- Sensor-driven quality data flowing directly into SPC charts — zero manual entry
Communication Protocols for the Shop Floor
Before connecting machines, you must understand how they communicate:
OPC-UA: The Modern Standard
OPC Unified Architecture is the gold standard for industrial interoperability. It provides platform-independent, secure communication with rich data modeling. Modern PLCs from Siemens, Beckhoff, Rockwell, and Mitsubishi support it natively. Use OPC-UA for direct PLC communication in multi-vendor environments.
MQTT: Lightweight and Versatile
MQTT is a publish-subscribe messaging protocol designed for constrained environments. It excels at sensor data transmission, edge-to-cloud communication, and high-volume data streams. Popular brokers include Mosquitto, HiveMQ, and AWS IoT Core.
Modbus: The Legacy Workhorse
Modbus TCP and RTU remain universally supported by older equipment. While lacking security features and event-driven capabilities, Modbus is essential for connecting legacy PLCs and instruments that predate modern protocols.
MTConnect: CNC-Specific
MTConnect is an open standard purpose-built for CNC machine monitoring. It provides a standardized data dictionary and RESTful HTTP interface, making it ideal for job shops with diverse CNC equipment.
Architecture Patterns for IoT Integration
Pattern 1: Edge Gateway Aggregation
This is the recommended approach for most facilities:
``` Machines (Modbus/OPC-UA) -> Edge Gateway -> MQTT -> Broker -> ERP API ```
Edge gateways normalize protocols, buffer data during network outages, and handle connectivity issues. Industrial options include Advantech, Moxa, and Siemens IOT2050. Cloud-managed options include AWS Greengrass and Azure IoT Edge.
Pattern 2: Direct Machine-to-Cloud
For newer machines with built-in MQTT or HTTP clients:
``` Modern Machine (MQTT native) -> Cloud Broker -> ERP API ```
Simpler but less resilient — no local buffering during network interruptions.
Pattern 3: MES as Middleware
When a Manufacturing Execution System is already deployed:
``` Machines -> MES -> ERP ```
The MES handles shop-floor orchestration and sends aggregated data to ERP. FlowSense includes a built-in lightweight MES module, combining both layers.
Data Model Design Principles
Universal Machine State Model
Define consistent states across all equipment:
- RUNNING — producing parts within specification
- IDLE — powered on, not producing
- DOWN_UNPLANNED — unexpected stoppage
- DOWN_PLANNED — scheduled maintenance or changeover
- SETUP — changeover between production runs
- OFFLINE — powered off or not communicating
Production Counting Best Practices
Use event-based counting rather than periodic polling:
- Total count — all parts produced including rejects
- Good count — parts passing quality criteria
- Reject count — parts failing, categorized by defect type
- Cycle time — actual time per part or per batch operation
Downtime Event Logging
Every state change generates a structured event with machine ID, timestamp, previous and new states, fault codes, and linked production order. The ERP processes these events to update order status, calculate OEE, and trigger alerts.
ERP Integration Points
Real-Time Production Tracking
IoT data feeds directly into the ERP production module:
- Work order progress updated automatically as parts are counted
- OEE calculated live with availability, performance, and quality components
- Schedule adherence visible in real time with early warning of delays
- Supervisor dashboards refreshed every 30 seconds
Automated Inventory Transactions
When production completes a step, the ERP automatically:
- Posts raw material consumption based on BOM and actual quantity
- Records WIP transfers between work centers
- Triggers finished goods receipt at last operation
- Posts scrap transactions with reason codes from quality sensors
Quality Data Integration
Inspection data flows from IoT devices to the ERP quality module:
- SPC charts generated from sensor data in real time
- Control limits monitored with automatic out-of-spec alerts
- Inspection records created without manual entry
- Non-conformance reports auto-generated when rejection thresholds are exceeded
Implementation Roadmap
Weeks 1-2: Audit and Plan
Inventory all machines, map available data points, prioritize by business value, and select edge gateway hardware. Document current network infrastructure and identify gaps.
Weeks 3-4: Pilot Line Setup
Install edge gateways on 3-5 pilot machines, configure protocol adapters, set up MQTT broker, define data model and topic hierarchy. Validate data flow end to end.
Weeks 5-8: ERP Integration
Build or configure REST API endpoints for receiving IoT data, implement real-time production tracking, configure automated inventory transactions, and set up downtime alerting.
Weeks 9-12: Validate and Scale
Compare IoT data against manual records, train operators on real-time dashboards, resolve edge cases, and begin rollout to remaining production lines.
Security Considerations
Shop floor IoT introduces new attack surfaces that must be addressed:
- Network segmentation — Keep OT network separate from IT with a DMZ
- Encryption in transit — TLS for MQTT, HTTPS for API calls
- Device authentication — X.509 certificates for edge gateways
- Firmware management — Establish a patch management process for IoT devices
- Anomaly monitoring — Detect unexpected traffic patterns indicating compromise
Cost and ROI for a 50-Machine Facility
| Component | One-Time Cost | Annual Cost |
|---|---|---|
| Edge gateways (10 units) | $15,000-$30,000 | $2,000 |
| MQTT broker infrastructure | $5,000-$10,000 | $1,500 |
| Network upgrades | $10,000-$20,000 | $3,000 |
| ERP integration | $20,000-$40,000 | $5,000 |
| Sensors and add-ons | $10,000-$25,000 | $2,000 |
| **Total** | **$60,000-$125,000** | **$13,500** |
Payback typically occurs within 8-14 months through improved OEE and reduced manual reporting effort.
Schedule a FlowSense IoT integration workshop to get a customized connectivity plan for your facility.
