This invention describes a system that uses artificial intelligence to find problems during manufacturing. It takes sensor data from factory equipment and specific production recipes, feeds it into a neural network, and then uses the network's output to identify unusual patterns or anomalies. The system can also use the specific recipe or equipment details to help the neural network process the data more effectively.
Why it matters: Filed in 2022, the intervening years have seen significant advancements in neural network architectures and training efficiency, particularly for industrial data. This makes implementing context-aware anomaly detection, as described, more robust and less resource-intensive to develop and deploy now.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI