In-Store Accessibility Navigation With Local Digital Twin Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing accessibility solutions for visually impaired individuals in retail environments face challenges such as slow AI responses, connection latency, reliance on unreliable cellular networks, and inconsistent human assistance, leading to frustrating shopping experiences.
Innovation Solution
A localized system that includes a digital twin of the store, leveraging local area networks for enhanced accessibility, utilizing a shopping list generator and navigation device to provide independent shopping assistance, and incorporating a verification process to ensure accurate product collection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing accessibility solutions (Seeing AI, Aira, Be My Eyes) are used, then accessibility for visually impaired individuals is provided, but response time is slow and connection latency is high
Solution Approach 1:
The patent extracts the accessibility service from centralized cloud-based systems and deploys it locally within the retail store environment. Local servers and edge computing nodes process visual information and generate navigation instructions without requiring continuous connection to remote servers, thereby eliminating connection latency and reducing response time while maintaining service reliability.
Solution Approach 2:
The patent introduces local intermediary servers and edge computing nodes between the visually impaired user's device and the centralized cloud. These intermediaries cache navigation data, process visual information locally, and relay instructions through the local area network, significantly reducing the effective response time and eliminating the need for continuous external connectivity.
2Reliability
If cellular networks are used for accessibility services, then connectivity is provided, but reliability is poor due to network instability
Solution Approach 1:
The patent employs local area networks as intermediary communication channels between user devices and the accessibility service. This local network infrastructure provides stable, low-latency communication that is independent of cellular network conditions, thereby eliminating the harmful effect of cellular network instability while maintaining reliable connectivity for accessibility services.
Solution Approach 2:
The patent transitions from centralized cloud-based services to locally-deployed accessibility infrastructure within the retail store. By placing servers and computing resources locally at the point of need, the system achieves higher reliability and consistency in service delivery, independent of external network conditions.
3Ease of operation
If human assistance is provided for shopping navigation, then guidance is available, but assistance is inconsistent and requires additional resources
Solution Approach 1:
The patent implements automated navigation guidance systems that independently provide turn-by-turn directions, product location information, and shopping assistance without requiring human intervention. The system uses computer vision, spatial mapping, and AI algorithms to autonomously navigate the store environment and guide visually impaired users, thereby ensuring consistent and reliable assistance availability.
Solution Approach 2:
The patent replaces the mechanical system of human assistance with an automated electronic navigation and guidance system. This substitution eliminates the variability and resource-dependency inherent in human-based assistance, providing consistent, reliable, and scalable support for visually impaired shoppers through software-based AI and computer vision technologies.
4Difficulty of detecting and measuring
If AI-based product identification is used, then product recognition is provided, but processing speed is slow
Solution Approach 1:
The patent segments the AI processing workload by deploying computer vision models and product recognition algorithms on edge computing devices and local servers within the retail store. This distributed processing architecture reduces the computational burden on any single system, enables parallel processing of multiple product identifications simultaneously, and significantly increases overall processing speed while maintaining accurate product recognition.
Data Source
AI summary
Architectures and techniques are described that can provide enhanced accessibility for in-store shoppers such as a shopper with a vision impairment. For example, the disclosed techniques can operate to dynamically generate a shopping list from freeform (e.g., speech) item descriptions. The shopping list derived from the item descriptions can include actual, specific product identifiers for products offered for sale at a physical store location. Furthermore, an associated in-store navigation route to products of the shopping list can be generated based on any one of several different collection techniques or approaches.