# How Training Systems Share AI Knowledge Without Sharing Private Student Data > This patent describes a system where different student training centers can improve their AI models by sharing general learning patterns, not private student data, to make training better for everyone. - **Patent:** US 11915111 - **Original title:** Federated machine learning in adaptive training systems - **Owner:** CAE - **Granted:** 2024 - **Status:** Active - **Times cited:** 1 - **Field:** software, ai_ml, telecommunications, aerospace, education, simulation ## What it does This system uses federated machine learning to improve student training across multiple locations while protecting privacy. A first training center uses an AI module to adapt training for its students and develops a 'first learning model' based on their performance metrics (Claim 1). A second training center extracts 'statistical properties' from its students' performance metrics, then uses a 'data simulator module' to create fake, but realistic, performance data (Claim 1). This simulated data is used to build a 'second learning model'. A central 'federation computing device' then combines the 'model weights' from both the first and second learning models to create or refine a 'federated model'. For example, in a flight simulator (Claim 8), pilot performance data like control yoke movements (Claim 8) could be used to improve the AI's ability to adapt training for new pilots at different flight schools. ## What it does NOT cover - Does not cover systems that share raw student performance data directly between training centers. - Does not cover machine learning systems that centralize all student performance data for model training. - Does not cover federated learning approaches that exchange full local models instead of just model weights. - Does not cover training systems where the AI does not adapt individualized training to each student. - Does not cover systems that don't use simulated data generated from statistical properties for a local model's contribution. ## The clever bit The clever part is how the system allows a training center to contribute to a shared AI model without revealing any actual student data. Instead of sending raw performance metrics, it extracts statistical properties and then generates simulated data, which is then used to train a local model whose 'weights' are shared. This multi-layered approach ensures privacy while still enabling collaborative learning for the AI. ## Real-world examples 1. Flight simulators for pilot training 2. Medical training simulations for surgeons 3. Military training systems for complex equipment operation 4. Industrial process control training platforms ## Why it matters This patent is significant because it addresses a major challenge in AI-driven education and training: how to leverage large datasets for model improvement while respecting data privacy. By using federated learning with simulated data, it allows multiple training centers to collaboratively enhance their adaptive training systems without directly sharing sensitive student information. This approach is particularly valuable for high-stakes training environments, such as aviation or medical simulation, where data privacy and robust AI models are critical. ## Frequently asked questions ### What does How Training Systems Share AI Knowledge Without Sharing Private Student Data cover? This patent describes a system where different student training centers can improve their AI models by sharing general learning patterns, not private student data, to make training better for everyone. ### Who owns patent US 11915111? CAE owns this patent, granted in 2024. ### When does this patent expire? This patent is expected to expire on March 15, 2043, when the invention enters the public domain. ### What is patent US 11915111 cited by? This patent has been cited by 1 later patents that build on its ideas. ### What problem does this patent solve? This patent is significant because it addresses a major challenge in AI-driven education and training: how to leverage large datasets for model improvement while respecting data privacy. By using federated learning with simulated data, it allows multiple training centers to collaboratively enhance their adaptive training systems without directly sharing sensitive student information. This approach is particularly valuable for high-stakes training environments, such as aviation or medical simulation, where data privacy and robust AI models are critical. ### What does this patent NOT cover? Does not cover systems that share raw student performance data directly between training centers. **Full plain-English explainer:** https://patentbrief.org/patent/us/11915111/federated-machine-learning-in-adaptive-training-systems **Original patent:** https://patents.google.com/patent/US11915111 --- _Source: PatentBrief — https://patentbrief.org. Patent facts are from public records; the plain-English explanation is PatentBrief's._ ## Related patents Semantically similar inventions in the PatentBrief corpus: - [Training AI on Private Data Without Seeing It](https://patentbrief.org/patent/us/12518214/distributed-machine-learning-systems-including-generation-of-synthetic-data) — This patent describes a way to train artificial intelligence models using private data stored on many separate computers, by generating fake data that mimics the real data's patterns, so the private data itself never leaves its original location. - [Training AI Models Together with Unlabeled Data Using a Teacher](https://patentbrief.org/patent/us/20220012637/federated-teacher-student-machine-learning) — This patent describes a way for multiple AI systems to learn together from data that hasn't been manually labeled, using a 'teacher' AI to create temporary labels for a 'student' AI. - [How Devices Train Shared AI Models While Keeping Your Data Private](https://patentbrief.org/patent/us/12443890/partially-local-federated-learning) — This patent describes a method for training a machine learning model across many devices, where each device keeps some parts of the model and its data private, only sharing updates for the common, global parts of the model. - [Sharing AI Knowledge Between Private Datasets Using Synthetic Data](https://patentbrief.org/patent/us/12039012/systems-and-methods-for-heterogeneous-federated-transfer-learning) — This patent describes a method for two separate AI systems to share learned information by generating artificial data based on common features, without directly exchanging their private training data. - [How Computers Train AI Models Using Separate Virtual Simulations](https://patentbrief.org/patent/us/11429762/simulation-orchestration-for-training-reinforcement-learning-models) — This patent describes a system where one virtual computer runs simulations of a system, like a robot, and another virtual computer uses the simulation data to teach an AI model how to make better decisions.