Imagine a camera that can see radiation, but instead of sending a massive, slow file to a computer to be processed, the camera chip thinks for itself. It uses a tiny built-in brain to instantly identify what it's seeing and only sends the important details. This makes radiation scanning incredibly fast and efficient, like switching from a slow dial-up connection to high-speed fiber.
Current radiation sensors are too slow and produce too much raw data, creating bottlenecks in real-time analysis for medical and security applications.
An AI-powered sensor readout system and a prototype test detector that uses embedded neural networks to classify radiation events instantly.
If you are a medical device company dealing with slow image processing and bulky data in nuclear medicine — this project developed an AI-integrated sensor chip that processes signals on-the-fly. This allows for faster, more accurate diagnostic imaging with significantly less data overhead.
If you are a space agency dealing with limited bandwidth and power for radiation monitoring in deep space — this project developed a system-in-package (SIP) that reduces data size from ~50 KB/event to ~16 bytes/event. This enables real-time monitoring without clogging communication links.
If you are a security firm dealing with the need for instant detection of hazardous materials at checkpoints — this project developed an intelligent readout system that performs real-time event classification. This ensures threats are identified with unprecedented speed and accuracy.
It drastically cuts digital data output, reducing the need for expensive high-bandwidth storage and processing infrastructure. For example, it can reduce event data size from ~50 KB to ~16 bytes.
The project focuses on a scalable sensor system-in-package (SIP) design, which is intended for integration into various hardware devices.
Based on available project data, the consortium includes 5 partners from 5 countries, but specific licensing terms are not detailed in the summary.
The project is active from 2024-03-01 to 2028-02-29, suggesting that commercial readiness would follow the 2028 conclusion.
The technology is designed as an 'all-in-one' SIP chip that connects directly to sensor signals for near-real-time processing.
The consortium is well-balanced for a technology transfer project, consisting of 5 partners across 5 countries (DE, DK, IT, NO, UK). With a 40% industry ratio (2 SMEs and 3 Universities), the project blends academic research in AI and physics with the practical manufacturing capabilities of SMEs, increasing the likelihood of a viable commercial product.
Contact DTU Space (Danmarks Tekniske Universitet)
Contact us to explore licensing opportunities for AI-driven radiation sensing.
