External Vision Control for Sensorless Mobile Robot Fleets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile robot control systems require onboard sensors and computers, leading to high costs, energy consumption, reliability issues, and operational inefficiencies due to sensor failures, especially in high-density environments.
Innovation Solution
A system using optical cameras and artificial intelligence processors to remotely control mobile robots without onboard sensors, utilizing high-resolution cameras, and deep neural networks for localization and trajectory control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If onboard sensors and computers are equipped on each mobile robot, then localization and control capability is improved, but system cost and complexity increase
Solution Approach 1:
The patent introduces external cameras and AI processors as intermediary components that mediate between the mobile robots and the control system. Instead of equipping each robot with complex onboard sensors and computers, the system uses external visual sensors to capture images of robots (with or without visual markers) and AI processors to analyze these images for localization and control, thereby reducing the complexity and cost of each individual robot while maintaining localization capability
Solution Approach 2:
The patent replaces traditional mechanical and electronic sensor systems (ultrasonic sensors, infrared sensors, electromagnetic sensors like RADAR or LIDAR, GPS) with an optical vision-based system. By using cameras to capture visual information and AI algorithms to process this information, the system substitutes complex mechanical and electronic sensing mechanisms with a simpler optical-based approach, reducing device complexity while achieving localization and control
2Measurement precision
If multiple sensors and computing systems are installed on each mobile robot, then localization and control capability is improved, but reliability decreases
Solution Approach 1:
The patent extracts the sensing and computing functions from the mobile robots themselves and places them in separate external systems. By removing onboard sensors and computers from each robot and consolidating these functions into external cameras and AI processors, the system reduces the number of potential failure points on each robot while maintaining the necessary localization and control capabilities
Solution Approach 2:
The patent merges the sensing and computing functions of multiple robots into a single centralized system. Instead of each robot having its own independent sensors and computer, the system combines these functions into shared external cameras and AI processors that serve all robots, thereby improving reliability through functional redundancy and centralized management
3Ease of operation
If onboard sensors and computers are equipped on each mobile robot, then autonomous navigation capability is improved, but energy consumption increases
Solution Approach 1:
The patent extracts the computationally intensive navigation and control calculations from the mobile robots and relocates them to external AI processors. By removing onboard computers from each robot, the system eliminates the continuous energy consumption required to power these computing devices on moving robots, while maintaining autonomous navigation capability through external image processing and control command generation
Solution Approach 2:
The system enables mobile robots to operate with minimal onboard components, relying on external infrastructure for complex processing. The robots themselves perform only simple execution of received control commands, while the energy-intensive tasks of localization, path planning, and collision avoidance are performed externally, allowing robots to conserve energy for their primary function of movement and task execution
Data Source
Figure 1~2
Figure 3~4
AI summary
METHOD AND APPARATUS FOR CONTROLLING MOBILE ROBOTS WITHOUT ONBOARD COMPUTERS OR SENSORS, comprising: one or more video cameras (C), preferably high-resolution and high-frame-rate, for capturing images of a spatial volume (S) located indoors or outdoors; processing units (3a, 3b, 3c), connected to the cameras (C), based on artificial intelligence technologies and deep neural networks for performing the individual localization and control of one or more mobile robots (R) without onboard computers or sensors, preferably including: an optical mobile robot detector (3a), a generator of individual identity and localization information (3b) for mobile robots (R) and obstacles (O), and a command processing unit (3c) necessary to control one or more mobile robots (R); a communication module (i) for sending digital messages containing the necessary commands (X) to control mobile robots (R).