Edge AI: Next-generation artificial intelligence

Whether it’s about automating processes or analyzing large data volumes, smart and self-learning systems are increasingly becoming an important part of business processes. In the past, these smart systems had to be connected to a cloud because that was the only way of providing the level of computing power required to run the mathematical algorithms. Now edge AI, short for edge artificial intelligence, is enabling the next generation of smart systems and transferring the know-how to the devices themselves.

Edge AI safeguards data sovereignty

Storing and evaluating data in clouds can lead to serious data protection problems. According to the Cloud Act, for example, US companies must release personal data when requested to by the US authorities—even if the data is stored on foreign servers. But edge AI can help ensure data sovereignty is retained in Europe by evaluating data directly where it is generated instead of in the cloud. This makes edge AI a matter of major strategic importance for Europe.


Edge AI enhances data economy

The Internet, web-enabled devices and data storage are extremely energy-intensive. But the majority of data that gets stored on servers is irrelevant. Edge AI paves the way for considerably lowering this energy consumption. Since only key metadata is sent to a server and sored there, edge AI could help reduce the amount of data sent to and stored on servers to a fraction of what it is


Edge AI shortens latency times

In terms of the speed at which data is processed, cloud-based artificial intelligence is also quickly approaching its limits – especially when it comes to advanced AI applications. Here too, edge AI promises a brighter future, characterized by faster data processing and greatly reduced latency times. In addition, devices will not require the stable network connection that cloud-based system


Edge AI: What’s involved?

Edge AI systems are made up of sensors, computing chips/processors that are used to carry out smart data analysis, and a communication interface that sends the metadata identified to the chosen server. Fraunhofer IIS is a capable contact for any questions relating to edge AI. The institute’s researchers develop, among other things, special neuromorphic hardware on which to perform energy-efficient AI calcula




Further information


AI series



Prof. Heuberger and Dr. Grill

AI series: Solutions that we cannot achieve with conventional methods


Prof. Alexander Martin

AI series: Mining data to extract knowledge


Dr. Norman Uhlmann

Near-sensor AI: Generating sensor data of outstanding quality


AI series - more stories


Your questions and suggestions are always welcome!

Please do not hesitate to email us.

Stay connected

The newsletter

Sign up for the the Fraunhofer IIS Magazine newsletter and get the hotlist topics in one email.

Home page

Back to home page of Fraunhofer IIS Magazine