Research Projects


Resource-efficient Thin-Edge systems through integrated AI and neuromorphic electronics in sensors

Start: 07/2022

End: 06/2025

The increasing number of connected devices and sensors, the “Internet of Things” (IoT), opens up a wide range of new application scenarios. It is also responsible for a rapidly growing data volume. Processing data at its point of origin (edge computing) supports efficient data handling. Edge computing strengthens the functionality, sustainability, trustworthiness, and cost-effectiveness of electronic applicationsby using Artificial Intelligence and networking.

As part of the ThinKIsense project, a predictive maintenance application will shift the data processing from the cloud directly to the sensor. Targeted data pre-processing in the sensor reduces the frequency of communication, which increases energy efficiency and, thus, the service life of the batteries in the sensor. Neural networks are used for pre-processing. Two techniques will be used to test acceleration: the first is through special instructions for artificial neural networks on a customized RISC-V system, and the second is through spiking neural networks and a neuromorphic accelerator matched with these networks.


Victor Pazmino Betancourt

Department Manager
Division: Embedded Systems and Sensors Engineering

Research Focus

Applied Artificial Intelligence

In this research focus of the FZI, the emphasis is on Human-centered Artificial Intelligence and AI engineering. The FZI also addresses questions on dedicated AI hardware and AI for prediction.

Climate Action Innovation

The research focus Climate Action Innovation aims to actively promote climate protection and provide sustainable solutions for energy, mobility, production, and supply and disposal via IT innovations. Security aspects of the solutions are considered from the very onset.

Funding notice:
The ThinKIsense project is funded by the Federal Ministry of Education and Research.

Project partners:

More projects