Robot Navigation Offloading for Lower Onboard Computing Load
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous robots face challenges in efficiently managing computing capacity, storage capacity, and power consumption for navigation tasks, particularly in determining behavior and generating control commands, which can lead to resource bottlenecks and software heterogeneity.
Innovation Solution
Outsourcing the computationally intensive determination of robot behavior to an external computing device, which processes sensor data, creates maps, and generates control commands for obstacle avoidance, thereby reducing the onboard computer's workload and enabling centralized navigation software updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If the robot's onboard computer performs behavior determination and control command generation, then the robot can operate autonomously, but the computing capacity, storage capacity, and power consumption increase
Solution Approach 1:
The navigation system is segmented into two parts: the robot's onboard computer handles real-time sensor data acquisition and basic navigation tasks, while an external computing device handles computationally intensive behavior determination and control command generation. This segmentation allows the robot to maintain autonomous navigation capability while reducing its own power consumption and hardware requirements.
Solution Approach 2:
An external computing device acts as an intermediary between the robot and the navigation software. The robot transmits sensor data to the external device, which processes the data and returns control commands. This intermediary approach enables complex autonomous navigation without burdening the robot's onboard computer with excessive computing and power demands.
2Extent of automation
If the robot's onboard computer performs behavior determination and control command generation, then the robot can operate autonomously, but the computing capacity and storage capacity increase
Solution Approach 1:
The computationally intensive functions of behavior determination and control command generation are extracted from the robot's onboard computer and relocated to an external computing device. This extraction allows the robot to maintain full autonomous navigation capability while significantly reducing the computing capacity and storage capacity requirements of its onboard computer.
Solution Approach 2:
The external computing device serves multiple robots simultaneously, providing centralized navigation software and processing capabilities. This multi-functional approach allows each robot to operate autonomously without requiring individual high-specification onboard computers, as they share the external computing resource.
3Extent of automation
If navigation software is distributed across multiple robots, then each robot can operate independently, but software heterogeneity and update management become problematic
Solution Approach 1:
The navigation software is merged into a centralized external computing device that serves multiple robots. This consolidation ensures that all robots run the same software version, eliminating heterogeneity issues and simplifying update management. Each robot operates independently by communicating with the centralized system, which handles software updates uniformly for all connected robots.
4Speed
If the robot processes all navigation data locally, then real-time response is achieved, but power consumption and computational load increase
Solution Approach 1:
The system dynamically distributes computational tasks between the robot's onboard computer and the external computing device based on task requirements. Real-time sensor data acquisition and basic navigation functions are handled locally for immediate response, while computationally intensive behavior determination is processed externally. This dynamic task distribution maintains real-time response capability while reducing the robot's power consumption.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
The invention relates to a method for operating an automatically moving robot (1). A map (2) of the surroundings of the robot (1) is generated using measurement data captured within the surroundings, and a control command is generated using the generated map (2), the current position of the robot (1) within the surroundings, and a determined behavior of the robot. The robot (1) is moved using the generated control command, and data which is relevant to the navigation of the robot (1) is at least partly transmitted to an external computing device (3) for processing. In order to reduce the computing capacity and/or storage capacity required within the robot (1), the external computing device (3) determines a desired behavior of the robot (1) as the basis for the control command based on the map (2) and the current position of the robot (1).