AI Personal Companion With Local-Global Models for Fast Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robotic companions lack advanced intelligence and personalization, limiting their ability to provide contextually relevant and autonomous assistance to users across different digital platforms and environments.
Innovation Solution
An artificial intelligence (AI) model is developed using a deep learning engine that integrates with a mobile platform, allowing the AI to access and classify user data, learn patterns, and build personalized models for autonomous companions, enabling them to interact with various digital assets and environments while providing contextually relevant assistance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If early robotic pets used simple processing power and restricted form factors, then they could move autonomously and interact with owners, but they lacked advanced intelligence and personalization capabilities
Solution Approach 1:
The system divides the AI model into two segments: a global AI model stored on a server that contains general knowledge and patterns, and a local AI model stored on the robotic companion that contains personalized user-specific patterns. This segmentation allows the robot to have advanced personalization capabilities while keeping local processing requirements manageable.
Solution Approach 2:
A server acts as an intermediary between the user and the robotic companion. The server stores and processes the global AI model, then transmits relevant patterns to the robot. This intermediary enables the robot to access advanced intelligence without requiring all processing power to be embedded in the device itself.
2Speed
If the AI model processes all requests locally, then response time is fast, but processing power requirements and device complexity increase
Solution Approach 1:
Instead of processing all requests locally, the system performs partial processing: the local AI model handles immediate responses using cached personalized patterns for fast response time, while more complex requests are processed by the global AI model on the server. This partial action approach achieves fast responses for common tasks without requiring excessive local processing power.
Solution Approach 2:
The system performs preliminary action by pre-processing user data and environmental information through the local AI model to generate personalized patterns, which are then cached for rapid retrieval. This preliminary processing enables fast response times during actual interactions without requiring the full processing power to be available during each interaction.
3Measurement precision
If the robotic companion collects and processes extensive user data, then personalization accuracy improves, but data privacy and security risks increase
Solution Approach 1:
The AI model is segmented into global and local components, with the local AI model on the robot handling only personalized user-specific data patterns, while the global AI model on the server handles general knowledge. This segmentation minimizes the amount of sensitive user data that needs to be stored and processed locally, reducing privacy and security risks while maintaining personalization accuracy.
Solution Approach 2:
The system extracts and separates sensitive personalization data from the robot device and stores it securely on the server in the global AI model. Only processed, less-sensitive patterns are transmitted to the robot. This extraction removes the harmful factor of data security risks from the device while preserving the beneficial effect of personalization accuracy.
Data Source
AI summary
A method for building an artificial intelligence (AI) model. The method includes accessing data related to monitored behavior of a user. The data is classified, wherein the classes include an objective data class identifying data relevant to a group of users including the user, and a subjective data class identifying data that is specific to the user. Objective data is accessed and relates to monitored behavior of a plurality of users including the user. The method includes providing as a first set of inputs into a deep learning engine performing AI the objective data and the subjective data of the user, and a plurality of objective data of the plurality of users. The method includes determining a plurality of learned patterns predicting user behavior when responding to the first set of inputs. The method includes building a local AI model of the user including the plurality of learned patterns.