ATM Interface Generation Using AI for Multilingual Accessibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ATM systems require manual and inaccurate processes for measuring dimensions and generating audio scripts, necessitating separate coding for sighted and visually impaired users, and lack support for multiple languages and dialects, failing to meet regulatory requirements.
Innovation Solution
A generative artificial intelligence-based system that uses AI models to analyze ATM dimensions, generate accurate audio scripts in multiple languages, and dynamically adjust user interfaces to meet regulatory standards and user preferences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual processes are used to measure ATM dimensions and generate audio scripts, then the process can be implemented with conventional systems, but the measurement accuracy and time efficiency deteriorate
Solution Approach 1:
The patent replaces manual mechanical measurement processes with an automated vision system using cameras and image processing algorithms. The system captures images of the ATM and automatically extracts dimensional information and component positions through computational analysis, eliminating the need for manual measuring tools and processes.
Solution Approach 2:
The system enables the ATM to automatically generate its own audio scripts and user interface descriptions without external assistance. The vision system captures images, processes them through AI models to extract spatial information, and automatically generates the corresponding audio narratives and UI element descriptions that the ATM can use independently.
2Reliability
If separate coding is done for sighted and visually impaired users, then specific accessibility requirements can be met, but the device complexity and development time increase
Solution Approach 1:
The patent implements a universal vision-based system that serves multiple functions: it generates audio scripts for visually impaired users, creates user interface element descriptions, and provides spatial information about ATM components. This single system replaces what would otherwise require separate coding efforts for different user groups, reducing overall system complexity while maintaining accessibility compliance.
Solution Approach 2:
The patent merges the generation of visual user interface elements and audio descriptions into a single integrated process. The vision system captures images that are processed to extract both visual UI information and spatial component positions simultaneously, generating both visual and audio outputs from one analysis pipeline, thereby reducing the need for separate coding streams.
3Adaptability or versatility
If conventional systems provide limited language options, then system complexity remains manageable, but adaptability to different languages and dialects deteriorates
Solution Approach 1:
The patent implements language adaptability by changing the language parameter of the generated audio scripts based on user preferences or detected demographics. The vision system and AI models generate content in a language-neutral format that can be translated or rendered in multiple languages, allowing the same system to serve diverse linguistic communities without requiring separate system implementations.
Data Source
AI summary
Arrangements for using generative AI models for ATM process generation are provided. A computing platform may receive user input from an ATM and may identify, based on the user input, a first generative AI model associated with a first functionality of the ATM. The platform may execute the first model to output a preferred language and a first plurality of options in the preferred language. The platform may generate and transmit, to the ATM, a first output providing the first plurality of options. In response, the ATM may receive second user input selecting an option of the first plurality of options. The platform may identify a second generative AI model associated with second functionality of the ATM. The platform may execute the second model to output a second plurality of options. The platform may generate and transmit, to the ATM, a second user output including the second plurality of options.