TOP

SBC - Edge Computing: 6 Uses Cases for Industrial IoT

Edge computing as holding a promise to no less than “dramatically” improve data processing for mission-critical applications and accelerate the adoption of Industry 4.0. The SBC-Edge Computing Platform, based on single-board computer and single-board edge clusters, provide an opportunity to extend the cloud computing paradigm to the edge where a SBC or micro cloud cluster exists between the data-center and the data source at the edge, which can significantly reduce the decision-making latency, save bandwidth resources, and to some extent, protect and privacy.

SBC-Edge Computing Platform for Industrial IoT

SBC - Edge Computing Platform for Industrial IoT 

 

The Challenge for Cloud Computing
In traditional cloud computing models, data needs to be transmitted over long distances to cloud service provider data centers for processing and analysis. This can result in latency and network congestion issues, which may affect application performance and efficiency. Additionally, transmitting large amounts of data can produce high costs.

 

What is Edge Computing

Edge computing is the deployment of computing and storage resources near the edge or source of data generation instead of sending all data to a centralized data center or cloud. By bringing computing resources and data storage closer to end devices and users, edge computing can avoid the latency and cost issues with data transmission. 

 

What is Edge Computing

What is Edge Computing

 

How Does Edge Computing Work?

Here is an example of how edge computing works:

 

Let's say you are using a smart thermostat in your home. Instead of sending temperature data to a cloud server for processing, the data is processed locally on the thermostat. The thermostat can then use this data to make adjustments to the temperature in your home in real-time.


To enable edge computing, specialized hardware, such as edge servers, edge routers, and edge gateways are required. These devices have enough processing power and memory to handle data processing tasks quickly and efficiently.

 

Single-Board Computers (SBCs) vs Commodity Computers for Edge Platforms

Edge platforms typically utilize commodity computers, which can be a significant source of energy consumption. These types of computers use approximately 200 W/h when being used, emitting around 525 kg of CO2 per year. Even in standby mode, power consumption remains relatively high. Data centers utilizing commodity computers consume roughly 1% of the world's electricity usage, which is considered a high percentage. In an effort to reduce energy consumption, Single-board Computers (SBCs) are becoming a popular replacement option for commodity computers. Because SBCs are smaller, less expensive, and consume significantly less energy - approximately 2.7 W in an idle state and 6.4 W in full load.

 

Single-Board Computers (SBCs) used as Egde Device

In edge computing, SBCs are often used as computing devices for edge nodes to process and analyze data in real-time, and can be used for tasks such as image processing, machine learning, and data analytics. Examples of SBCs include Raspberry Pi, Arduino, BeagleBone and LattePanda. In addition, data privacy is ensured by edge computing since data are being processed locally instead of at the cloud. Yet, with data becoming a higher liability and powerful single board computers (SBC) decreasing in cost, the best solution is often to bring at least some processing back to the local network.

Basic Architecture of Edge Computing

Basic Architecture of Edge Computing

 

Edge Computing is Essential for Industrial IoT (IIOT)

The term IoT, Internet of Things, is often used to refer to ubiquitous, consumer-oriented IoT products. The Industrial Internet of Things (IIoT) , on the other hand, stands for Industrial Internet of Things, and is a subset of the larger IoT, focusing on the specialized requirements of industrial applications such as manufacturing, oil and gas, and utilities.

 

IIoT requires both cloud-based data centers and edge processing to function efficiently. While cloud-based resources offer mass storage, processing power, and security, they can also experience latency issues and increased exposure to hackers. To combat these issues, processing and analytics can be performed at the edge with the help of single-board computers (SBCs). 

The IoT and IIoT Application Domains

The IoT and IIoT Application Domains 


Use Case Examples for IIoT Edge Computing with Single-Board Computers (SBCs)     

  • Use Case: NVIDIA Metropolis for Factories

Many manufacturers have automated optical inspection (AOI) systems that can help, but often these have high false detection rates, requiring labor-intensive and costly secondary manual inspections in an already challenging labor market, reducing their value.

 

NVIDIA Metropolis for Factories now offers a state-of-the-art AI platform and workflows for the development of incredibly accurate inspection applications such as AOI. Its customer using the entire Metropolis for Factories workflow to support its printed circuit board (PCB) factories with simulation, robotics and automated production inspection. Metropolis for Factories enables the electronics manufacturing giant to quickly update its defect detection models and achieve 99.8% accuracy on its AOI systems, starting with small datasets.

PCB Defect Detection with NVIDIA's Metropolis for Factories

PCB Defect Detection with NVIDIA's Metropolis for Factories

 

The platform is built on the foundations of NVIDIA’s Nova Orin reference architecture, which serves as the brains and eyes of Isaac AMR. Nova Orin integrates multiple sensors, including stereo cameras, fisheye cameras, and 2D and 3D lidars, with NVIDIA’s Jetson AGX Orin system-on-modules. 

 

Why NVIDIA’s Jetson AGX Orin? 

  • Up to 248 TOPS of AI performance with power configurable between 15-75 W. 
  • Supports multiple concurrent AI application pipelines with an NVIDIA Ampere architecture GPU
  • Deep learning and vision accelerators, high-speed I/O, and fast memory bandwidth 
NVIDIA’s Jetson AGX Orin

NVIDIA’s Jetson AGX Orin

 

  • Use Case: Warehouse Visualization: 

A visualized warehouse requires a data warehousing system and needs the installation of sensors to collect real-time data. These sensors can be installed on goods, shelves, trucks, and even on staff to collect important data such as temperature, humidity, transport location, motion status and more during the logistics process. The data is transmitted to a Single-board Computer like LattePanda 3 Delta, and after analysis and processing, can be displayed on the visualized warehouse interface, allowing warehouse managers to easily understand inventory status for daily management and decision-making.

LattePanda 3 Delta Powered Warehouse Visualization 

 

In a smart warehouse, in order to monitor the worker's location, posture, and movement, identifying potential safety issues or hazards and sending alerts to the worker or warehouse manager in a timely manner, the sensors are installed on the staff helmet, timely translate date to SBC for date processing.

 

Why choose LattePanda 3 Delta? 

  • Strong Computing Power: CPU: Intel Celeron N5105
  • Memory: LPDDR4 8GB 2933MHz & Storage: 64GB eMMC V5.1
  • Ultra-fast connectivity: Wi-Fi 6 and Gigabit Ethernet
  • Higher data transfer rate: 1 x USB3.2 gen2, 2x USB 3.2 gen1, and 1x USB Type C for high data transfer speed, up to 10X faster than USB2.0
  • Comparability: Support Windows 10, Windows 11 and Linux
Single-Board Computer (SBC): LattePanda 3 Delta

Single-Board Computer: LattePanda 3 Delta 

 

  • Use Case: Smart-parking Transformation with ASUS IoT Tinker Board S

The smart-parking transformation with ASUS IoT Tinker Board S contains the intelligent-vehicle-access-control system (IVACS) and a payment kiosk(EYEPAY), creating a comprehensive, cloud-connected parking-control solution that operates autonomously while being permanently connected to remote-management administration systems. This industry-first arrangement enables data to be stored securely in the cloud, eliminating the risk of loss due to failure at the local site. It also ensures both IT and management personnel are able to monitor parking resources and payments in real time.

Smart-parking transformation with ASUS IoT Tinker Board S

Smart-parking transformation with ASUS IoT Tinker Board S

 

Why choose ASUS IoT Tinker Board S?

  • Rockchip RK3288 Quad Core ARM Cortex-A17 Processor
  • ARM-based Mali-T764 GPU, 4K Resolution
  • 2GB LPDDR3 Memory, 16GB eMMC Flash Storage
  • 40-Pin GPIO Interface, MIPI DSI/CSI Support
  • SBC Dimensions: 3.37" x 2.125"
ASUS IoT Tinker Board S

ASUS IoT Tinker Board S

 

  • User Case: Automation and Control System

Raspberry Pi Compute Module 4 can be used in Automation and Control System to perform a variety of computing tasks such as data acquisition and processing, control and monitoring of processes, and communication with other devices. It can also provide a platform for developing and running custom software applications tailored to specific automation and control tasks. Additionally, it also can be integrated with various sensors, actuators, and outputs to form a complete automation and control system.

 

Why choose Raspberry Pi Compute Module 4 (CM4)?

  • Two rows of high-speed, high-density mezzanine connectors
  • Broadcom BCM2711 Quad-Core Cortex-A72 (ARMv8)
  • A variety of RAM and eMMC storage configurations
  • Small size: 55mm × 40mm
Raspberry Pi Compute Module 4

Raspberry Pi Compute Module 4

 

  • Smart Manufacturing with SBC-edge Computing for Deep Learning

The use of smart manufacturing in the IIoT has several advantages. New industrialists are adopting IIoT solutions to enhance the productivity and profitability of their industries. The integration of DL methods in smart industries can upgrade the smart manufacturing process into a highly optimized environment by information processing through its multi-layer architecture. Deep Learning(DL) inference and training require substantial computation resources to run quickly. Edge computing, where a fine mesh of compute nodes are placed close to end devices, is a viable way to meet the high computation and low-latency requirements of deep learning on edge devices and also provides additional benefits in terms of privacy, bandwidth efficiency, and scalability. 

Some well-known Development Boards for DL Implementations

Some well-known Development Boards for DL Implementations

 

Use Case Examples for IIoT Edge Computing with SBC-edge clusters (SBC-EC)

  • Use Case: Parking Application based on SBC-edge clusters (SBC-EC)

Due to SBC’s small form factor and low costs, SBC based edge clusters can be easily and cost effectively deployed on a site. The sensors data can be relayed to the SBC-edge clusters (SBC-EC) for local processing instead of the distantly located data-center, therefore reducing the need for massive network communication.

 

The SBC-EC’s ecosystem leverages the use of light weight cloud computing technologies suitable for resource constraint SBC devices. The framework utilizes Docker based containers, Containers are designed to be lightweight having a small foot print in comparison to the traditional virtual machines (VM). Much of the operating system functions such as file system, memory address space and networking can be virtualized in a container. The small size of containers as opposed to a VM makes it easier to copy and move the container images between devices.

 SBC- EC framework

 SBC- EC framework

 

The smart city parking application using the SBC- EC framework. The application deployment on the cluster serves as a prototype implementation of the framework. The main objective of this application is to log data collected from parking sensors, process the data locally in the SBC-EC (total occupation time of parking bays) and send updates to the cloud service.

Raspberry Pi, Odroid Xu-4, LatttePanda V1

(a)Raspberry Pi  (b)Odroid Xu-4  (c) LatttePanda V1

 

Conclusion
The combination of SBC and edge computing provides a powerful platform for Industrial IoT. SBCs offer high performance and reliability, making them suitable for edge computing. They can also collaborate with the cloud for more flexible and efficient IoT applications. Using SBCs as edge computing platforms in industrial IoT deep learning enables real-time data processing and analysis, improving production efficiency, quality control, and predictive maintenance. However, challenges remain in selecting the appropriate SBC platform based on factors such as application requirements, reliability, budget, software support, and community ecosystems. Despite these challenges, the combination of SBCs and edge computing has significant potential in industrial IoT deep learning.

 

Reference: 1. Deep Learning for the Industrial Internet of Things (IIoT): A Comprehensive Survey of Techniques, Implementation Frameworks, Potential Applications, and Future Directions