zh 
Using edge AI edge, you can bring computing capacity directly to the source, thereby reducing latency and the energy spent on moving most of the processing to the cloud, as well as improving data security. Taken separately, edge computing and Artificial Intelligence (AI) are already capable of providing significant benefits to a business, but combined they can truly revolutionize business strategy, helping to cut overall costs and improve performance, which are two benefits that it’s really difficult to overlook.
The greater availability on the market of SFF (Small Form Factor) electronic components and systems equipped with low energy consumption, but increasingly powerful processors (CPU), has made it possible to decentralize calculations, so that the ability to analyze data provided by Artificial Intelligence can be performed “locally”, directly on devices and sensors. This significantly reduces the amount of data that needs to be sent to the cloud for subsequent analysis and sharing, reducing the costs of particularly onerous activities, such as centralized processing.
How edge AI can be more convenient than the cloud
Therefore, edge AI represents the optimal solution for all cases in which the amount of data that needs to be sent and subsequently processed in the cloud leads to high transfer costs and delayed response times. For example, video surveillance cameras in airports, which have to process thousands of images every day. The combination of edge devices and Artificial Intelligence allows data processing directly on board the cameras, avoiding unnecessary bottlenecks during the phase of sending mass data to a central server for analysis. Analyzes that, requiring significant computing capacity, might not offer the required performance in order to, for example, activate control and facial recognition, rescue, safety and maintenance services.
Therefore, time is a decisive factor when choosing the most appropriate infrastructure for the production of certain goods or services. Time that, in many cases, is measured in terms of the latency (delay) with which data captured is processed and passed on to a "decision maker", whether human or digital, such as another application. For example, in the automotive sector or that of high-speed rail transport, where hundreds of sensors need to communicate the data captured in just a few milliseconds, in order to avoid accidents. Low-latency transmission not only means greater safety in mission-critical applications, but also becomes an economic benefit, for example when applied to the issue of maintenance.
The quality of the infrastructure through which data transmission takes place also affects latency: if the bandwidth is insufficient or the connection is not stable or secure, then the delay increases. The arrival of 5G technology in the sphere of the IoT and in edge AI, more specifically, can further boost this sector, guaranteeing higher speeds, a greater number of simultaneous connections and even lower latencies. All of the above will have positive repercussions on all fields that use edge AI, from Automotive to Industrial IoT, with great benefits for the entire Industry 4.0.
Edge AI: a better return on investment
Therefore, the interaction between edge computing and AI clears the path for new data-driven approaches, improving the reliability of industrial processes and enhancing products and services. It is no coincidence that, in May 2020, the IBM Institute for Business Value calculated that the expected ROI from investments in edge solutions for the three-year period 2019-2022 is well above 5% on average for all the main sectors, with peaks nearing 9% for the automotive sector, 7% for industrial and retail products and 6% for transport and public administration (smart cities).
In short, it’s a revolution with a strong impact on business models thanks to improved scalability and opportunities for customization, significant reductions in energy consumption and greater savings on the costs of data analysis. All of the above despite the initial investment required for hardware and software components and any staff training required on the use of the new platforms. The ability to have better quality data in a shorter timeframe has a positive effect on all the fundamental processes in a company, from assistance and maintenance to product compliance, warranties, sales and customer relations.
Now, more than ever, the right choice of partner to help the company along this new path is essential to rapidly implement these significant technological innovations. Innovations that not only lead to greater performance and quality of services, but also to renewed data reliability as a support to all the company’s business goals.
