Standing on the shoulders of ESH, playing edge computing is not a dream

Not long ago, Rackspace, a well-known cloud service provider, surveyed 1,870 IT executives from all walks of life around the world to understand the progress of enterprises in AI and machine learning. The results show that the biggest difficulty hindering companies from implementing AI is the lack and difficulty in recruiting corresponding talents.

At the same time, the Internet of Things and edge computing are becoming important positions for AI applications. The IoT research group under the well-known Eclipse Foundation in the IT industry found that artificial intelligence is the most common of edge computing workloads, accounting for 30%. In the next 5 years, machine learning will promote the AI ​​IoT market to reach a compound annual growth rate of 27.3%.

From this, if you can master the development of AI edge computing, both the future and the money will be bright.

However, the development of edge computing involves countless hardware combinations, as well as a variety of software, algorithms, and models. At the same time, complicated and unpredictable network structures must be considered. As an ordinary developer, it is really “a tiger eats the sky and has nothing to say”. This also makes it difficult for companies’ smart edge solutions to go on the market in time, thus losing the opportunity to compete in the market.

Not long ago, Intel released the Intel Edge Software Hub (hereinafter referred to as ESH). It is a one-stop resource platform. Developers can choose software packages for specific vertical industries, customize the configuration according to their needs, and then determine the target hardware for deployment, thereby shortening the solution development process from several weeks to several hours. Realize prototyping and product integration, saving development time and development costs.

Standing on the shoulders of ESH, playing edge computing is not a dream

From hardware products, development tools to integrated software stacks

Intel provides complete development resources

The software tools and development kits provided by ESH have been pre-tested and verified to meet commercial deployment requirements. They integrate various edge software stacks, including the visual edge insight platform, the converged edge insight platform, the industrial edge insight platform, and the industrial edge control platform. . Using these edge software resources and reference architectures, developers can focus on the realization of the business layer and application layer, avoiding a lot of low-level development and architecture design work. At the same time, Intel provides a wide range of scalable and compatible hardware products, coupled with a rich industry ecosystem, including support from open source communities, ecological partners, and cloud service vendors, all of which make edge computing easier .

Starting from the case, development is faster and easier

Anyone with software development experience knows that if you want to get started quickly with any technology, detailed steps and a wide range of reference cases are indispensable. This is even more true for a large and complex technology field like edge computing. In view of this, the Intel Edge Software Center provides a wealth of reference use cases. They are built on the edge software stack, integrate the Open Model Zoo deep learning model and other third-party software modules, and can fully implement various specific edge AI applications. Developers can not only learn from it, but also develop secondary development based on these reference use cases to quickly build edge AI application prototypes and solutions.

Let us look at two examples.

01. Intelligent traffic management:

Open the use case, you can download the corresponding software components, including: Docker CE, Docker Compose and OpenVINO Container, and then directly run the sample code, you can realize the detection, tracking, collision detection, interest area, data storage and visualization in intelligent transportation, etc. Features.

Standing on the shoulders of ESH, playing edge computing is not a dream

02. Multi-camera social distance detection:

This use case can detect personnel, calculate the social distance between personnel based on multiple input sources, mark and count offenders, and visualize it in real-time in the browser. The feature detection model is used in the use case. After running some simple configurations, you can see the results shown in the following figure: the blue box indicates the safety distance, and the red box indicates that the safety distance is violated.

It is worth pointing out that all software resources, including edge software stacks for various vertical industries and reference use cases for various practical application scenarios, can easily download and install scripts with a simple operation of a web browser. Using the command line interface to execute the script, you can automatically install and deploy the software. After the installation is complete, you can continue to use the command line to query, update, and uninstall software resources. This makes the development experience easier and faster. Of course, supporting documentation and introductory manuals are indispensable; at the same time, there is also a dedicated ESH technical support forum in the Intel Developer Zone.

Easily obtain software resources through the Edge Software Center

Create the future, join us

With the Intel Edge Software Center, the whimsical ideas of countless small and medium-sized enterprises and third-party developers will no longer be difficult to land due to limited resources. With the help of ESH, these ideas have space for Display, which not only gives enterprises and individuals the opportunity to flourish, but also benefits the whole society.

If you also have a pair of eyes that are good at discovering opportunities and a heart that is uncomfortable with the ordinary, you may as well join the Intel Edge Software Center and work with us to realize the ideal of technological transformation of the world.