SANOVO TECHNOLOGY GROUP has signed a worldwide exclusive license agreement with JigLabs to further develop optical-based computer vision system, utilizing artificial intelligence for future implementation in crack detection systems.
For many years automatic egg crack detection has been based on acoustical measurements. Sound is generated by the egg hitting sensors or small hammers hitting the egg, creating a vibration of the eggshell, which is transferred to a sensor. This sensor converts the vibration into an electrical signal, analyzing and detecting egg cracks.
Sound-based systems are common and standard in crack detection. However, those approaches require that eggs have physical contact with the sensors/hammers. Not only does it expose mechanical stress on the eggshell, but it is also increasing the risk of cross-contamination from one egg to another.
Sound-based crack detection systems, especially with moving hammers, are not designed for food- grade contact with safe and easy cleaning. They require manual cleaning and occasionally dismantling for inspection. A much more sanitary design solution are the systems where eggs hit fixed sensors and often, cleaning of these sensors is designed to be part of the overall system through automated CIP (cleaning-in-place). Nonetheless, there still is a physical touch between egg and sensor that is posing a potential risk of cross-contamination.
With increasing food safety concerns, this area calls for innovative crack detection solutions. Therefore, SANOVO can proudly announce having signed an agreement with JigLabs, a software development company with founders from US and Europe. JigLabs provides an optical-based computer vision system that utilizes deep learning for artificial intelligence to detect cracks and leaks in an egg.
Jan Holm Holst, R&D Director at SANOVO: “Optical detection of egg cracks, especially hair-line cracks, have been tried for many years and several patents have been applied for that, but a successful operational system has never been obtained. For this reason, the sonic-based crack detection systems stayed in the market despite an increase of using computer vision technology in more advanced applications. I must say that this new and special designed artificial neural network
algorithm, imbedded in advanced, super high performing hardware, has made a revolutionary change to the future of egg crack detection, and I am sure that in due time we will see this technology taking over all the old sonic based systems used today.”
Until now, tests have been conducted in the US on white eggs. The test site has seen immediate improvements with the new SANOVO deep learning-based computer vision detection system that they decided to turn off their existing, old sound-based crack detectors.
- More stable detection results can be achieved, due to the elimination of mechanical moving parts and no impacts/pollution on sensors
- Almost free-of-maintenance since the vision system is based on modern camera technology, and light sources to illuminate the eggs are LED-based
Jan-Willem Pennings, R&D Manager at SANOVO: “When I was introduced to the concept, I saw all the potential benefits for the owner/operator of such a system. The crack detection concept is totally non-contact-designed, compared to the available sonic systems in the market. This means no potential further damage to the final product. The system desires no extra transfers with e.g. grippers or change of a carrying surface. Again, less damage through transfer of eggs. The system is static, so no moving parts, resulting in low maintenance and ultimately operating cost for our customers. In addition to that, no further adjustments regarding egg size are needed, like for sonic systems. I am also happy to see that the first test results show great promises in both accuracy and performance, and I cannot wait to bring it to the market to revolutionize crack detection for our customers.”
To turn this innovation into a solution that can be used for the detection of all types of eggs, SANOVO and JigLabs are currently working on system integrations and optimization of the technology. With high expectations and excitement for the future, we are looking forward to revolutionizing crack detection in egg graders all around the world.