Shelf Inventory Robot With Multi-Angle Vision and Edge AI
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing inventory systems struggle to accurately count items in enclosed spaces where camera views are obstructed, leading to inefficiencies and high error rates due to obstructed views of products on shelves.
Innovation Solution
A mobile mechanical device with movable appendages equipped with cameras and additional sensory modalities, such as microphones and touch sensors, is used to position itself for multiple perspectives, combined with edge-AI computing to dynamically retrieve context-specific models for accurate item counting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a static camera or mobile camera is used to take inventory in an enclosed space, then the camera can capture images of visible objects, but the view is obstructed by objects in the foreground, preventing accurate counting of all items on shelves
Solution Approach 1:
The patent introduces a movable appendage that can rotate about its longitudinal axis and move along vertical and horizontal dimensions. This multi-dimensional movement capability allows the camera to change its viewing angle and position, capturing images of objects that would otherwise be obscured from a single fixed viewpoint, thereby resolving the view obstruction problem.
Solution Approach 2:
The appendage is designed with dynamic movement capabilities including rotation about the longitudinal axis and translation along vertical and horizontal directions. This dynamic positioning system enables the camera to adaptively adjust its viewpoint to capture obscured objects, transforming the static imaging problem into a dynamic multi-perspective solution.
2Productivity
If manual inventory counting is performed using mobile phone applications, then the process can be automated to some extent, but it still requires manual operation to cover physical premises, making it time-consuming and costly
Solution Approach 1:
The system employs an autonomous mobile robot that independently navigates through the enclosed space, positions its appendage to capture images of objects, and processes the images using AI algorithms. This self-service capability eliminates the need for manual operation, allowing the robot to continuously perform inventory counting without human intervention, thereby dramatically improving productivity and reducing time loss.
Solution Approach 2:
The patent replaces manual mechanical counting operations with an automated robotic system equipped with computer vision and AI processing. The robot autonomously performs navigation, image capture, and object identification, substituting human labor with an integrated mechanical-intelligent system that operates faster and more efficiently.
3Measurement precision
If cameras operating in visible or ultraviolet electromagnetic spectrum are used, then they can capture images of objects, but non-transparent objects block the camera view, preventing complete inventory assessment
Solution Approach 1:
By enabling the camera to move in multiple dimensions through the movable appendage, the system captures objects from various angles and positions. This multi-dimensional imaging approach allows the camera to bypass electromagnetic blockage by finding alternative viewing paths around obstructing objects, thereby maintaining object detection capability despite spectrum limitations.
Data Source
AI summary
An apparatus for automating inventory procedures for items stored on shelves in a closed environment includes a mobile mechanical device having a movable appendage including a camera and additional sensory modalities. The camera and additional sensory modalities are used to position the movable appendage to take camera images of the items on the shelves from many different perspectives. Camera vision and the additional sensory modalities can be used to rotate, lower and raise movable appendage to position the appendage over and along the sides of the items on the shelf. The context of the mobile mechanical device is determined and edge AI computing retrieves AI context specific models based on the context. The AI context specific models may be downloaded from a cloud service. The edge AI computing uses the AI context specific models to identify and count the items on the shelf.