An edge computing appliance performs resource-intensive processes or analysis at or near the source of the data. It solves the need for IoT and IIoT devices to be continually sending data to the cloud or remote data centers for further processing and analysis.
An edge computing appliance brings the power of the cloud to you.
The NCR-1510 and LEC-2290, are two of the latest platforms added in Lanner’s wide range of intelligent edge computing product line. NCR-1510 is a rackmount rugged edge network appliance, suitable for wide-temperature applications, While LEC-2290 is a high-performance A.I. accelerated industrial edge server designed for image processing.
What is an Edge Computing Appliance?
An edge computing appliance is simply a device that processes data from a massive number of devices or software, on-site.
But an edge computing appliance is NOT the ordinary local computer or mobile device doing limited computational tasks on edge— instead, it is an edge platform that brings the computing and storage power, only found in the cloud or high-speed data centers. Wherever your edge network is, at a remote site, at a branch office, or a headquarters data center, an edge computing appliance will bring the cloud’s intelligence closer to the edge, whether it is AI/ML analytics, mobile computing, cloudlets (offloads processing tasks), even storage.
An edge computing appliance reduces or eliminates (in some cases) the need to send data to an offshore server. Leaving data on-site, provides fantastic benefits such as:
- Reduce application and network latency exponentially.
- Improved bandwidth.
- Transferring less data (or no data) over the Internet increases security.
Depending less on third-party networks, storage, and computing, improves reliability and control.
Lanner’s Edge Computing Appliance Solutions
An edge computing appliance is critical for optimizing IoT and IIoT devices and other applications by bringing the computing capacity closer to data sources. This edge computing hardware is a type of edge technology that aims to reduce the need for long-distance communication between the server and the client. Ultimately decreasing latency, and improving security, reliability, and bandwidth.
Lanner’s edge computers portfolio showcases appliances with multiple functionalities. These appliances are not limited to general edge computing capabilities, but they are optimized for wide-temperature scenarios, rugged environments, industrial edge AI, and computer vision applications.
LEC-2290: The Industrial Edge Box
Lanner’s LEC-2290 is a high-performance industrial edge computing appliance designed for compute-intensive workloads. This edge platform runs demanding AI-powered video analytics and other AI/ML algorithms at the edge. With its rich I/O and its powerful GPU capabilities, LEC-2290 is king in image processing at the edge.
An essential distinction of LEC-2290 is that its appliance’s computing capacity can be expanded with GPUs, VPU, and FPGA for additional video processing power and AI applications.
- Support for Intel® Core™ i7-8700T/i7-8700.
- Rich I/O with four PoE. 6x RJ45 GbE LAN ports, 6x serial ports, 8x Digital DI and 8x DO, 4x USB Type A, Ethernet, and WiFi/LTE expansion. And high power PoE for network apps.
- GPU Expansion. LEC-2290 can be expanded with GPU, VPU, and FPGA for powerful video processing.
- Optional wireless expandability. 1x PCIe x16, 1x Mini PCIe w/ Nano-SIM, 1x B Key M.2 w/ Nano-SIM
Additionally, LEC-2290 supports Intel’s OpenVINO toolkit and Movidius interface engine. The OpenVINO toolkit from Intel (Open Visual Inference and Neural network Optimization toolkit) provides developers with neural network performance to help create real-time computer vision applications.
The Movidius interface, also from Intel, is a Vision Processing Unit (VPU) that helps accelerate the system to perform resource-intensive computer vision and edge AI algorithms.
NCR-1510: The Rackmount Rugged Network Appliance
Lanner’s NCR-1510 is a fanless network edge server with a tolerance for wide temperatures. The device is optimized for SD-WAN and/or uCPE. With a rugged edge computing appliance, service providers can deploy edge intelligence in demanding environments. The NCR-1510 comes in a hardened white-box which can be used in mobile edge cell tower-based applications, energy plants, manufacturing floors, farms, oil rigs, or any environment exposed to wide temperatures.
- Wide operating temperature ranging from -40 to 70°C.
- Powerful Processing. Intel Atom® 2~8 Cores C3308/C3508/C3708 CPU (Denverton) with QuickAssist Technology.
- Rich I/O. 4x RJ45, 2x SFP or 2x RJ45 (By SKU), 1 Pair of Gen3 Bypass.
- Optional Wireless Connectivity. 2x Mini PCIe for Wi-Fi, LTE (signal: USB2.0, PCIe) with Nano-SIM.
What are the Applications of an Edge Computing Appliance?
With edge technology, you can run Artificial Intelligence (AI) and Machine Learning (ML) workloads closer to the edge, without having to send massive quantities of data to the cloud.
The edge technology of these appliances makes them an excellent fit for the following applications.
- IoT edge. Edge computing was created due to the massive growth in IoT devices. These devices would connect to the Internet to send or receive data from the cloud. Sometimes they would need to transfer massive amounts of data while operating. So the IoT edge is the concept of using an IoT gateway to enable users to perform edge computing with their IoT devices. The edge computing appliance may serve as the gateway and/or as the data processing unit.
- Industry 4.0 and IIoT Edge. An edge computing appliance brings IIoT devices together. For example, IIoT sensors across a conveyor belt on a manufacturing floor might feed data (wirelessly or wired) to the edge compute hardware on-site. Then, the edge appliance runs AI analytics, any ML algorithms, or any kind of data processing to gain insights, without needing to send data to the cloud.
Examples of some Edge Computing Appliance use cases
As mentioned before, the IIoT edge is the combination of Industrial IoT devices, their data, and the edge intelligence applied to that data. The IIoT edge can be applied to the smart factories, including one of the most popular use cases: manufacturing plants. When edge computing is combined with Private 5G, the smart factory can have all of its IIoT sensors connected to an edge computing appliance. The appliance can transfer IIoT data to other branches or headquarters using the Private 5G wireless.
COVID-19 Prevention. Having computer vision capabilities at the edge (especially on main entrances) may help to prevent COVID-19 infections. For example, a set of cameras, a high-performance edge computing appliance, and computer vision algorithms will help enforce COVID-19 compliance through video intelligence. A solution like this can detect people without a mask, promote social distancing through crowd detection, create watchlists (with face recognition) of infected people, and automatically detect fever with thermal imaging at entry points.
Some of the qualities of a “Smart City” are intelligent systems of IoT sensors, networks, video surveillance, and more. These systems may use edge computing to provide faster response and more security on the city’s applications. Real case scenarios of how the edge computing appliance can be used in a smart city, may range from: detecting traffic anomalies, bad drivers, unusual crowds of people, recognize criminals through face, Intelligent Transportation Systems, water treatment, garbage management, and more.
The edge computing appliance may also benefit retail stores. The store’s endpoints from POS terminals, inventory server, payment controllers, etc. can be virtualized and centrally managed (remotely or locally) from the edge computing server — this is referred to as the Software-Defined Retail Store. Running a compute-intensive workload consolidation appliance, such as the LEC-2290, guarantees many VMs to run simultaneously. It also provides an opportunity to connect may store the device’s peripherals.
Edge computing helps by processing data gathered from farms or agricultural environments without a fast connection to the cloud. Farms are usually rural environments, which can be harsh for deploying the technology. An edge computing appliance such as NCR-1510, can operate on extreme temperature conditions and provide compute-intensive workloads processing at the edge. Real use cases of edge intelligence in rural areas may include precision agriculture, drone surveillance or rural video surveillance, water flow monitoring, etc.
Micro-Data Centers and Mobile Edge Computer
Micro-Data Centers (MDC) are sometimes used interchangeably with edge computing. But in fact, MDCs are just small and modular data-center architecture boxes built to work almost in any environment. An edge computing appliance may run as a core computing element in an MDC. The Mobile Edge Computer (MEC), is the same as MDC but aimed for mobile networks. A MEC processes, stores, streams, and provides security at the edge of mobile networks, usually in the RAN (Radio Access Networks) instead of the core.