Edge computing has become a popular concept. To that end, Unisinsight’s EC201 edge computing node is an ideal solution in any project where edge processing is needed. This article takes a closer look, based on product testing by a&s China.
Edge computing has become a popular concept. To that end, Unisinsight’s EC201 edge computing node is an ideal solution in any project where edge processing is needed. This article takes a closer look, based on product testing by a&s China.
Edge computing is gaining importance in video surveillance. Yet there are various challenges associated with edge computing. Cameras with intelligent edge processing capabilities are oftentimes expensive. Further, backend equipment may also need to be upgraded. All this leads to rising project costs.
In response to this, Unisinsight recently launched their edge computing node EC201mini-I (hereafter referred to as “EC201”). EC201 combines IoT terminal access, video management, storage, retrieval, and intelligence features all into one and is supported with high computing power and reliability. It also features a variety of advanced deep learning algorithms which can perform multi-target and multi-behavior analysis. With multi-terminal connection and deployment flexibility, EC201 can meet the intelligent upgrade needs by various industries.
Flexibility and reliability
EC201 has different ports to connect a range of devices. There is an LED status indicator on the front, with which the user can directly check the operating status of the node. The back is equipped with ports for: networking, video, audio, USB2.0, USB3.0, alarm, M.2, RS232/485 and CAN2.0B, making EC201 deployable in a variety of scenarios.
It is worth noting that with the M.2 interface, the EC201 can have its functionality expanded to include SSD, 4G/5G, Wi-Fi, accelerator card and other capabilities, meeting even more user needs and requirements.
On the inside, the EC201 is equipped with 8GB LPDDR4x memory, 32GB eMMC flash memory, and a built-in 128GB TF card. The overall power consumption is less than 20W, which effectively saves energy and meets the current sustainability requirements. The device has a size of only 190×126×38mm and weight of 0.9Kg. It can work normally in an environment with a temperature range of -40-70℃ and humidity of 80-90 percent. The power supply voltage is only 12V. And thanks to the IP40 protection rating, EC201 can be placed in any outdoor equipment box and still work normally.
In addition, EC201 is extremely simple to deploy. The user only needs to connect it to a power source and the Internet, and EC201 will automatically register to the system and start doing what it’s supposed to do. This helps reduce the installation cost, speed up project deployment, reduce maintenance work and raise overall work efficiency.
IoT and AI empowerment
As an edge computing node, EC201 supports Unisinsight’s own communications protocol as well as cameras based on ONVIF, RTSP, and GB/T28181 protocols. In a&s’s test, it was found the EC201 has an access/storage/conversion rate of 80Mbps and can accommodate up to 10 cameras with 1080P@30fps resolution.
It is worth noting that EC201 has a built-in Unisinsight algorithm, which can perform structured analysis on people, motor vehicles, and non-motor vehicles in the video and extract corresponding feature values. It can support databases of up to 1.5 million vehicles and 1.5 million non-motor vehicles.
Further, image-based smart search can be performed. With the front-end camera, applications such as people flow, crowd density and perimeter detection are also enabled. In addition, users can choose from various algorithms, such as smart mines, smart water conservancy and kitchen brightness, to fully meet their personalized needs.
It is worth noting that EC201 also supports multi-algorithm fusion and can run multiple intelligent analysis algorithms at the same time to meet project demand.
Although Unisinsight’s self-developed algorithms can meet users’ various needs, their demand and requirements can change from time to time. To this end, Unisinsight has built an open AI algorithm library to effectively integrate the algorithms from industry partners. To make third-party algorithms easier to integrate with EC201, Unisinsight has simplified access to the open AI algorithm library. Algorithm providers only need simple development interfacing to be able to have their algorithms run on EC201. At the same time, algorithm upgrade, install and uninstall will not affect the normal operations of EC201.
To better meet users’ algorithm capability requirements, EC201 can be configured for 7Tops, 16Tops, and 32Tops computing power, and the M.2 interface can also be used as a computing power card slot to further enhance the computing power of EC201.
In addition, EC201 also supports multi-protocol access expansion, such as Modbus, OPC-UA, BACNet and MQTT, as well as the expansion of third-party industry protocols to allow connection by different IoT devices. This will further facilitate IoT project deployment.
Features:
- Supports international and Unisinsight protocols and auto-registration to the system;
- Supports cameras based on ONVIF, international and Unisinsight’s own protocol;
- Can perform structured analysis on up to 10-channel video or picture streams;
- Supports the storage, retrieval and playback of smart, regular and alarm-based recording;
- Software-defined hardware effectively meets the specific needs of various industries;
- Algorithm upgrade, install and uninstall will not affect audio/video operations;
- Supports access expansion of multi-protocols, such as Modbus, OPC-UA, BACNet and MQTT, to meet various user requirements.
a&s review:
With the development of IoT, the front-end capturing devices encompass not only cameras, but also various sensing devices. This will lead to increased equipment purchases, maintenance cost and overall complexity of projects. Against this backdrop, a node device that ingests and analyzes data at the edge becomes increasingly needed.
As an edge computing node, EC201 can be effectively deployed in any environment due to its small size and low power consumption. Based on its computing power and stable algorithm, EC201 can perform structured analysis on video captured by cameras and extract relevant feature values to improve overall work efficiency. At the same time, EC201 also supports Modbus, OPC-UA, BACNet and other protocols, effectively meeting users’ need to connect multiple IoT devices to the EC201.
Further, Unisinsight provides not only general AI algorithms but also industry-specific advanced AI algorithms to achieve in-depth applications in specific scenarios. To better meet the personalized needs of users, Unisinsight has built an algorithm library to effectively integrate the algorithm capabilities of industry partners. This effectively allows a fusion of algorithms and avoids algorithms influencing or interfering with each other.
Based on the characteristics of EC201, it can be widely used in new/existing projects for the following fields: parks, finance, transportation, construction, general education, cultural tourism, water conservancy, energy and others.