In an industrial IoT system, the data acquisition gateway acts as the "sensor tentacle"—it connects sensors, instruments, and PLCs in the physical world, converting various forms of signals into standardized digital data for upper-layer platforms to analyze and utilize. However, industrial field signals vary greatly: continuously changing temperatures, on/off statuses, messages from serial ports, and data packets transmitted over networks.
Understanding the collection methods and data types of an industrial data acquisition gateway is the first step to correct selection and deployment. This article systematically explains these two core issues, from technical principles to practical applications.
Industrial data acquisition gateways obtain data from field devices through four main methods, each corresponding to different signal types and interfaces.
Technical principle: Analog signals are continuously changing physical quantities, such as temperature, pressure, flow rate, liquid level, current, and voltage. Sensors convert these physical quantities into standard analog electrical signals (commonly 4-20mA current or 0-10V voltage). The gateway's analog input (AI) channels sample at high resolution (e.g., 12-bit, 16-bit), converting the continuous analog quantity into discrete digital values.
Typical applications:
Temperature transmitter (4-20mA corresponds to 0-100°C) → Gateway collects mA value → Converts to temperature value
Pressure sensor (0-10V corresponds to 0-1.6MPa) → Gateway collects voltage → Converts to pressure value
Key parameters:
Sampling accuracy (e.g., ±0.1% FS)
Sampling frequency (e.g., once per second, 100 times per second)
Number of input channels (4, 8, 16)
Technical principle: Digital signals have only two states: 0 or 1 (on/off, high/low). They include two categories:
Discrete input: Relay contacts, button status, valve position signals, motor running status.
Pulse input: Flow meter pulse output, encoder counts, electricity meter pulses.
The gateway's digital input (DI) channels detect high/low levels (e.g., dry contacts, wet contacts at 24V), recording status changes or pulse counts.
Typical applications:
Monitoring pump running status (DI=1 means running, DI=0 means stopped)
Measuring conveyor belt speed (calculated from pulse frequency)
Detecting limit switch activation
Key parameters:
Input type (dry contact / wet contact)
Isolation voltage (to prevent strong interference)
Maximum counting frequency (e.g., 10kHz)
Technical principle: Many industrial devices (PLCs, smart meters, VFDs, barcode scanners) have serial communication interfaces (RS232, RS485, RS422) and output structured data via protocols such as Modbus RTU, Profibus DP, and DL/T645. The gateway acts as a master, polling slave devices by sending read commands; the devices respond by returning data frames.
Typical applications:
Reading voltage, current, and power from multiple electricity meters via RS485 bus (Modbus RTU protocol)
Connecting to a PLC via RS232 to read its internal registers (e.g., temperature, production count, fault codes)
Connecting to a weighbridge indicator to obtain weight data
Key parameters:
Number of serial ports (1-4)
Baud rate (9600~115200bps)
Supported protocol library (Modbus, DL/T645, CANopen, etc.)
Technical principle: Modern industrial devices commonly have Ethernet ports or wireless network cards, supporting TCP/IP, Modbus TCP, Profinet, EtherNet/IP, OPC UA, and other industrial Ethernet protocols. The gateway establishes connections with devices via network cables or Wi-Fi/4G/5G, directly reading their data points or variables.
Typical applications:
Collecting PLC data blocks (DB blocks) via Modbus TCP
Connecting to CNC machines via OPC UA to read spindle load and feed rate
Receiving JSON data from sensors via MQTT
Key parameters:
Number of Ethernet ports and speed (10/100/1000Mbps)
Supported protocol stack (TCP, UDP, MQTT, OPC UA)
Maximum number of simultaneous connections
Through the four collection methods above, the gateway can obtain the following four types of data:
| Data Type | Raw Form | After Gateway Processing | Typical Application |
|---|---|---|---|
| Analog signal data | 4-20mA current, 0-10V voltage | Temperature 35.6°C, Pressure 0.85MPa | Process control, environmental monitoring |
| Digital signal data | Dry contact on/off, TTL level | Motor status = Running, Count = 1250 | Equipment start/stop monitoring, pulse accumulation |
| Serial communication data | Modbus RTU message (hexadecimal) | Three-phase voltage, current, power factor | Smart meter reading, PLC data acquisition |
| Network communication data | TCP/IP packets, JSON strings | Equipment parameters, alarms, production cycle time | Remote monitoring, MES integration |
This is the most basic data type for gateways. After conversion by sensors and transmitters, physical quantities like temperature, pressure, and flow become standard electrical signals. The gateway collects them and, using built-in linear transformation formulas (e.g., y = kx + b), converts them back to engineering units.
Value: Real-time awareness of key process parameters for closed-loop control or limit alarms.
Digital signals tell managers "yes or no," "present or absent." They do not contain complex numerical values but are essential for equipment status monitoring.
Value: Determining whether equipment is running, faulty, or in position; counting start/stop cycles and runtime.
Compared to analog and digital signals, serial communication can transmit richer information. A single Modbus read command can retrieve dozens of parameters (voltage, current, power, frequency, temperature, etc.) with high accuracy and strong noise immunity.
Value: Achieving multi-parameter, multi-device centralized acquisition at lower cost; the most mainstream method in industrial sites.
Network communication supports large data volumes and high-frequency acquisition. For example, OPC UA can read hundreds of variables per second; MQTT can receive JSON data packets pushed by edge devices.
Value: Suitable for smart manufacturing, digital twins, and other scenarios requiring massive real-time data.
In actual projects, select a gateway with the appropriate collection methods based on the signal types and communication capabilities of field devices.
| Device Type | Common Signal/Interface | Recommended Collection Method |
|---|---|---|
| Temperature sensor (without transmitter) | Thermocouple, RTD (requires transmitter) | Analog (after transmitter) |
| Pressure/level transmitter | 4-20mA / 0-10V | Analog |
| Limit switch, relay | Dry contact | Digital |
| Smart meter, flow meter | RS485 + Modbus | Serial communication |
| PLC (Siemens, Mitsubishi, etc.) | Ethernet + proprietary protocol | Network communication |
| VFD, servo drive | Profibus, CANopen | Serial communication (requires protocol conversion) |
How many AI channels are needed? Accuracy requirements? Sampling frequency?
How many DI channels are needed? Is counting functionality required?
How many serial ports are needed? Do they support field device protocols (Modbus, DL/T645, etc.)?
Are Ethernet ports needed? Which industrial Ethernet protocols are supported (Modbus TCP, Profinet, OPC UA)?
Many sites have multiple signal types simultaneously. For example, a pumping station needs:
AI: Collect pipe pressure (4-20mA)
DI: Collect pump running status (dry contact)
Serial: Read electricity meter data (Modbus RTU)
Ethernet: Upload data to cloud platform (MQTT)
In such cases, choose a hybrid collection multi-function gateway to avoid deploying multiple dedicated devices.
Fact: The serial port is only the physical layer. The protocol layer (e.g., Modbus, Profibus) determines whether communication is possible. Confirm whether the gateway's protocol library includes the field device's protocol.
Fact: Sensor accuracy, signal cable shielding, grounding, and gateway sampling accuracy together determine final data quality. Using shielded twisted pair cable with single-ended grounding is important.
Fact: Network communication is suitable for large data volumes and high frequency, but it is more complex to configure and more expensive. For Modbus RTU acquisition with a few dozen points, a serial solution is more economical and stable.
The collection methods and data types of an industrial data acquisition gateway determine what an IoT project can "see" and to what extent it can "control." Correctly identifying the types of signals on site (analog/digital/serial/network) and selecting a gateway with the corresponding collection capabilities is the foundation of project success.
In actual deployment, it is recommended to first use a multimeter, serial debugging tool, etc., to confirm field signal characteristics, then match them against the gateway datasheet. When the gateway's indicator lights flash steadily and data is accurately presented on the platform, this "first mile" is truly opened.