EV Charger Visual Guidance Using LLMs for Session Troubleshooting

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Electric vehicle drivers face challenges in navigating inconsistent charging processes due to variations in charger and vehicle hardware, firmware, and software configurations, leading to difficulties in accessing relevant information and troubleshooting charging issues.

Innovation Solution

Systems and methods utilizing visual context and large language models (LLMs) to analyze user interactions and charger environments, generating context-specific prompts for LLMs to provide real-time guidance and control commands for charging sessions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If visual data collection and LLM processing are implemented to provide context-specific charging guidance, then charging session success rate and user guidance quality improve, but system complexity and computational resource requirements increase

Engineering Contradiction:

Improvecharging session success rateVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces an LLM-based intermediary system that acts as a mediator between the charging infrastructure and users. This intermediary processes visual data from cameras, interprets charging session contexts, and generates appropriate guidance or control commands, thereby improving charging success rates while managing system complexity through intelligent abstraction

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces traditional mechanical or rule-based charging control systems with an AI-driven LLM system. Instead of relying on pre-programmed sequences and hard-coded responses, the system uses natural language processing and visual recognition to dynamically generate charging guidance, improving adaptability and success rates

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If LLM-based real-time analysis and visual processing are used to identify charging issues, then troubleshooting accuracy and support response time improve, but computational energy consumption and processing time increase

Engineering Contradiction:

Improvetroubleshooting accuracyVSAvoidcomputational energy consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent implements preliminary action by pre-processing visual data and preparing context information before LLM analysis. The system captures and pre-analyzes charging session data, vehicle information, and charger status in advance, so that when troubleshooting is needed, the LLM receives pre-organized information, reducing real-time computational energy requirements while maintaining high troubleshooting accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent applies partial action by selectively processing only the most relevant visual and contextual data for each charging issue. Rather than analyzing all possible parameters continuously, the system focuses computational resources on identifying and resolving the specific problem at hand, reducing overall energy consumption while maintaining high diagnostic accuracy

Inventive Principle:

Principle #16Partial or excessive action

3Adaptability or versatility

If comprehensive visual data processing and multi-source information integration are implemented, then user guidance relevance and context awareness improve, but information processing time and computational resources increase

Engineering Contradiction:

Improvecontext awarenessVSAvoidinformation processing time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent extracts and separates critical information elements from the comprehensive visual and contextual data. By identifying and extracting only the most relevant features (such as charger status, vehicle type, connection state) before LLM processing, the system reduces information processing time while maintaining high context awareness and guidance relevance

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20250381876A1Systems and methods for facilitating charging sessions between electric vehicles and chargers

Publication Date: 2025.12.18 CHARGEMATE AI INC

AI summary

A system for charging a target vehicle using a target charger is provided. The system comprises the target charger and one or more cameras positioned in a vicinity of the target charger. The system determines a hardware type and a software version of the target charger; obtains, from the one or more cameras, visual data corresponding to a physical environment of the target charger; identifies one or more features including the target vehicle, the target charger, and a user; and generates a system prompt for a large language model (LLM) based on the hardware type and/or the software version of the target charger, and/or the one or more features. The system transmits the system prompt to the LLM; receives an output from the LLM; and charges, or displays an instruction to the user to charge, the target vehicle using the target charger in accordance with the output from the LLM.