No matter what kind of IoT system is built, it must have three "must-have" elements: sensor nodes, gateways and cloud networks. The gateway is a transitional system between the sensor node and the cloud, responsible for the transmission of information and commands in both directions. However, due to distance limitations, gateway functionality has been improved, and now shares part of the processing workload with the cloud, which is now considered an edge computing resource.
Many different types of devices can be considered gateways in different IoT systems. For example, in home appliance IoT applications, smart speakers may act as gateways. Each node connects to the speaker via Wi-Fi or Bluetooth, which transmits information and commands. Another example might be a public-facing IoT application. The cellular base station acts as a gateway, and nodes are connected to them via 4G LTE or 5G, and information and commands are transmitted by the cellular base station.
gateway is also used in industrial Internet of Things (IIoT) applications that require industrial computers (IPCs). In IIoT, there are many examples of gateways. Applications may simply use embedded computers as gateways, while other applications use full-size server rooms. However, many gateways used in the Industrial Internet of Things can only perform basic functions and lack comprehensive software and hardware support.
Recently, I received Advantech EI-52 (EI-52-S6A1) edge intelligent system for experience. I unboxed it, tested the system, and found it was easy to use. In this article, I will use EI-52 to illustrate the key points and features that should be considered when setting up a gateway for an IIoT application.
EI-52
Three applications for industrial-grade computers
IPC is very different from ordinary home and office computers. They may be installed next to high-temperature boilers or in areas with high chance of collision in hot spots of visitors. Therefore, the mechanical design of industrial control machines needs to pay more attention to reliability. The mechanical design of the EI-52 is not only designed for extreme environments, but also has an industrial-grade rugged design that can operate at temperatures ranging from -10 to 50 °C.
EI-52 has three main applications; first, indoor facilities, such as shops, shopping malls and department stores; second, factory production line automation; third, self-service kiosks for retail and convenience stores.
Mainstream Windows operating system with long-term services and support
After investigating the server market in the past, I noticed that tower servers are the most widely used servers in the retail industry. Even with the growing market share of Linux, Windows operating systems remain the mainstream operating system used in the retail industry, as store managers and employees are more familiar with the operating system, and they can more easily change settings or perform basic troubleshooting under Windows. EI-52 Using Windows 10 Enterprise Long-Term Servicing Channel (LTSC), EI-52 provides not only a familiar operating system, but also guarantees long-term after-sales service and support.
EI-52 Use Windows 10 Enterprise Edition Long-term Service Channel
EdgeX Foundry and a powerful technology ecosystem
EdgeX Foundry is a highly flexible and scalable open source software framework that promotes interoperability between IoT edge devices and applications. In practice, the gateway needs to use Docker container technology to execute EdgeXFoundry.
Due to the major upgrade of EdgeX Foundry version 2018, its security infrastructure has been enhanced and computing resource demand has been reduced. Furthermore, it now uses Docker container technology, rather than Hypervisor (VM, virtual machine), is lighter, retaining a large amount of hardware resources for other purposes and future expansion.
In terms of lower resource requirements, the old version of EdgeX, including containers, requires 519MB of storage space, while the new version only requires 68MB. When simulating 100 nodes running, the old version requires 1302MB of RAM, while the new version only requires 27MB. In addition to this significant savings, gateway service programs are initialized faster, from about 35 seconds per service to less than a second. I checked the storage space used by all pre-installed software in EI-52.It only has about 38 GB, including the operating system, EdgeX Foundry and all related software programs.
EdgeX Foundry has a strong technology ecosystem thanks to the support of a variety of hardware and software vendors. Its products include a variety of built-in communication protocols, drivers and sample code for multiple sensor nodes such as cameras and barcode readers. The relevant documents, historical issues, and discussion records are also very comprehensive. EdgeX Foundry truly simplifies and speeds up the development of IIoT applications.
EI-52's EdgeX control panel interface
WISE-DeviceOn, an efficient back-end management software
In terms of operation, EI-52 uses EdgeX Foundry, which uses microservices to manage nodes. In addition, gateways also need management, so Advantech developed a device management solution called WISE-DeviceOn.
WISE-DeviceOn is divided into two parts. One is WISE-Agent, which is installed on the managed device (EI-52 here), and the other is WISE-DeviceOn cloud service, established and operated by Advantech as the backend. WISE-Agent is responsible for collecting information about managed devices and transmitting them to the backend, while receiving commands to control devices from the cloud.
Through the backend, administrators can view the real-time status of multiple gateways on site, such as network connection status, exception events, etc. If an abnormal event is detected, further diagnosis can be performed remotely. In addition, the equipment data can be further analyzed for predictive maintenance.
Administrators can wirelessly update firmware on all gateways, remotely control power switches, and configure notification conditions. More importantly, WISE-DeviceOn allows grouping of multiple gateways (device groups) so that batch operations can be performed. This saves a lot of time and effort.
WISE-DeviceOn cloud management interface (several covers of device names)
Public cloud support
EI-52 After collecting all information on the site, it needs to be uploaded to the cloud. In theory, any cloud service can be used as long as the data format and protocol are correctly mapped to the required cloud service standards.
As mentioned above, the cloud is an indispensable part of the entire Internet of Things operation. Through data aggregation in the cloud, administrators can understand the overall changes in field devices, discover subtle patterns from collected data analysis, and make subsequent predictions and responses.
EI-52 Online Microsoft Azure IoT Hub's backend management screen (several covers some information)
Two optional functions: inference acceleration and backup
EI-52 can perform AI inference locally because it shares inference workload with the cloud, which makes it more timely, power-saving, and reduces data transmission costs (public cloud services usually charge by data usage).
For this, the 11th generation Intel® Core™ i5/i3/Celeron processors in EI-52 use the built-in AVX-512 VNNI instruction set that can be calculated using hardware. You can also choose to add Advantech's VEGA-330 AI module to the EI-52, which is packaged in mini-PCIe and includes two Intel® Movidius™ Myriad™ X VPUs to speed up inference in video streams.
In addition to the Edge AI module, the EI-52 can also be equipped with Advantech other 5G and Wi-Fi modules to meet various communication needs of the cloud and sensors. All modules are tested for compatibility, performance and thermal reliability to ensure stable operation.
Gateway has higher requirements for information security than sensor nodes and cloud networks, because most cloud servers are located in data centers, the system is subject to strict monitoring and multi-layer protection, while gateways and sensor nodes are mostly placed in open environments and public spaces, which makes it difficult to have sufficient protection. Gateways can not only be hacked through the Internet, but also physically.
When a single sensor is attacked, it usually only causes small local damage, and the workload is easily overwhelmed by adjacent nodes. However, once the gateway is attacked, the impact will be extended to the entire area and cannot be ignored.
Therefore, optional antivirus and data backup and recovery software utilities are installed in EI-52. Advantech has partnered with software vendors to test and verify that McAfee antivirus software and Acronis data backup and recovery software can be used to provide all necessary protection and security measures for EI-52.
In addition to software utility, EI-52 also has a hardware chip (Infineon SLB9670XQ2.0), which provides TPM 2.0 security-related functions to avoid data tampering. At the same time, data is saved in a decentralized manner, and even if a hacker accesses the entire EI-52 device, the data in it cannot be correctly decoded without the verification key.
EI-52 solution overall schematic diagram (drawing: Lu Xiangyang)
Conclusion
From this unboxing experience, we can conclude that although general simple and basic gateway solutions provide sufficient functions, a comprehensive edge intelligent solution is necessary to operate stably and reliably for a long-term, stable and reliable manner. Such solutions should have robust mechanical design, good after-sales service, a strong software developer ecosystem, remote diagnostics/firmware updates/batch operations, inference acceleration, and data security and protection. Focusing on these key points, EI-52 is an ideal edge computing product that plays an active role in IIoT applications.
