Chapter 3: Industrial Internet of Things Perception

Objectives of This Chapter

1. Understand the connotation of the perception layer in the Industrial Internet of Things.
2. Learn about the current status of multi-source heterogeneous devices and systems in industrial scenarios.
3. Understand the scope, characteristics, and architecture of industrial data collection.
4. Learn about the categories and forms of device access products.
5. Understand the principles of common physical sensors.
6. Understand key indicators of industrial data collection. Familiarize with wired and wireless device access.
7. Understand short-range wireless communication and applicable scenarios.
8. Understand long-range wireless communication and applicable scenarios.
9. Understand the protocol conversion process.

Key Terms

Perception layer, industrial data collection, physical sensors, multi-source heterogeneous devices and systems, fieldbus, industrial Ethernet, industrial control networks, enterprise information networks, non-real-time data, real-time data, protocol conversion, industrial gateways, wireless device access, Wi-Fi, low-power Bluetooth (BLE), radio frequency identification technology (RFID), cellular networks.

1. Connotation of the Perception Layer

The perception layer serves as a bridge between the physical world and the digital world, acting as the first entry point for data. In reality, due to the diverse sources of data from various multi-source heterogeneous devices and systems, how to obtain data from these devices and systems is the first hurdle faced by the Industrial Internet of Things.

In the industrial field, perception is commonly referred to as industrial data collection.

The perception layer is located at the bottom of the Industrial Internet of Things architecture, responsible for “identifying objects, collecting information, and automatic control,” serving as the entry point for data interaction with the physical world.

2. Current Status of Multi-source Heterogeneous Devices and Systems in Industrial Scenarios

In industrial sites, various devices and systems coexist, including sensors, actuators, PLCs, SCADA, ERP/MES, etc. The multitude of communication protocols (fieldbus, industrial Ethernet, cellular networks, etc.) creates a “multi-source heterogeneous” pattern, posing challenges for unified access.

3. Scope, Characteristics, and Architecture of Industrial Data Collection

• Scope: It includes data from industrial site devices, as well as data from smart products, environments, and personnel outside the factory, and even data from traditional information systems.

• Characteristics: Multiple industrial protocols coexist, data often carries timestamps and has high real-time requirements, with a short “shelf life” that necessitates quick processing.

• Architecture: Typically layered as “sensor-gateway-edge-cloud,” where local format conversion and preliminary processing are completed before uploading to the cloud for further analysis.

4. Categories and Forms of Device Access Products

• General controllers: Equipped with rich I/O and various communication ports, can directly connect to fieldbus/Ethernet.

• Dedicated data acquisition modules: Optimized for specific signals (e.g., temperature, pressure), compact in size, and flexible in installation.

• Smart products and terminals: Equipped with processors and communication modules, capable of making local decisions and actively reporting data.

5. Overview of Common Physical Sensor Principles

• Temperature and humidity sensors: Convert environmental temperature and humidity into electrical signals using thermistors or capacitive humidity elements.

• Pressure/displacement/acceleration sensors: Utilize piezoresistive, capacitive, or piezoelectric effects to convert mechanical quantities into electrical quantities.

• Acoustic/flow/photoelectric sensors: Detect sound waves, fluid velocity, or light intensity and output corresponding electrical signals.

Sensor outputs can be analog or digital and need to be matched according to the input range of the acquisition card or MCU.

6. Key Indicators of Industrial Data Collection and Wired/Wireless Access

• Key indicators: Sampling rate, accuracy, resolution, linearity, temperature drift, anti-interference capability.

• Wired access: Primarily fieldbus, industrial Ethernet, and standard Ethernet, offering high reliability and low latency, suitable for fixed devices.

• Wireless access: Cellular networks, LPWAN, Wi-Fi, Bluetooth, etc., with low wiring costs and convenient for monitoring mobile assets.

7. Short-range Wireless Communication and Scenarios

• Technologies: Zigbee, Z-Wave, Bluetooth, Wi-Fi local networks, etc.

• Typical scenarios: On-site equipment fault diagnosis, personnel positioning, short-term data backhaul, requiring low power consumption and self-organizing capabilities.

8. Long-range Wireless Communication and Scenarios

• Technologies: Cellular 2G/3G/4G/5G, LPWAN (NB-IoT, LoRaWAN).

• Typical scenarios: Cross-plant pipeline monitoring, logistics tracking, urban-level environmental monitoring, emphasizing wide coverage, large connection numbers, or high reliability with low latency.

9. Protocol Conversion Process

The protocol conversion occurs at the gateway or edge node, completing four steps: “protocol identification → message parsing → data mapping → packaging → forwarding”:

• Identification: Automatically match or manually load the protocol library based on device model.

• Parsing: Split the original frame into fields and verify CRC/ASDU.

• Mapping: Map fields to a unified data model (e.g., OPC UA, MQTT).

• Forwarding: Package according to the target protocol and send to the cloud or MES/ERP.

This process achieves “multi-protocol coexistence and unified data view,” which is key to bridging OT and IT.