Toradex launches Klepsydra’s Edge AI demo in the Torizon Demo Gallery
Zug/Horw - Development teams can now explore the first Edge AI demo in the Torizon Demo Gallery from Toradex. Klepsydra, a Zug-based company that specializes in embedded computing, is behind the demo, which allows for experimentation with established deep learning models.
The Zug-based software firm Klepsydra has announced, that its Edge AI demo has been launched in the Torizon Demo Gallery. This is the first Edge Artificial Intelligence (AI) demo on the platform, which is provided by Toradex. It allows developers to experience AI workflows directly on supported embedded hardware and explore the possibilities of deploying deep learning models in edge environments.
Toradex, which operates globally out Horw in the Swiss canton of Lucerne, specializes in embedded hardware and software solutions. Its Torizon Demo Gallery offers various demo applications that enable developers to quickly assess their performance and various frameworks, without having to build anything from scratch.
"We are excited to bring the first Edge AI demo to the Torizon Demo Gallery", comments Klaus Buchheim, Head of Business Development at Klepsydra."Supporting ONNX gives developers a straightforward way to experiment with established deep learning models and understand how they can be deployed in real-world edge environments."
The Klepsydra demo runs on Apalis and Aquila boards from Toradex and demonstrates how modern Edge AI workloads can be brought closer to where data is generated. This approach facilitates responsive, efficient and locally executed AI applications without having to rely solely on cloud-based processing.
Klepsydra develops Edge AI software for the space, robotics and automotive sectors. Founded as a spin-off from the California-based robotics company Farm-ng in 2018, the company offers patented Edge AI software solutions that enable computers to process up to ten times more data while consuming 50 percent less energy. Through its Pattern project that received validation earlier this year, Klepsydra has extended the compatibility of its Edge AI to key European and international space-grade processor architectures. This facilitates real-time decision-making on board by enabling satellites to process data locally. Thereby, insights for Earth observation, autonomous navigation and collision avoidance can be generated far more quickly, all of which serves to reduce dependency on ground stations.
