Adaptive Robot Navigation for High-Quality Sensing Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques for Artificial Intelligence (AI) enabled adaptive navigation for autonomous roaming robotic devices, particularly for optimizing sensing capabilities, face challenges such as poor quality data due to environmental dynamics, sensor positioning errors, and variability in operational environments.
Innovation Solution
A computer-implemented method and system that utilize AI to optimize sensing capabilities of a roaming robotic device. This involves receiving data from the robotic device, analyzing it to determine quality, generating a model based on sample data, comparing input data to the model, and dynamically adjusting the robotic device's navigation to improve sensing quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a robotic device physically navigates to collect data in dynamic environments, then mobility and versatility are improved, but positioning and orientation errors increase leading to poor data quality
Solution Approach 1:
The system continuously receives sensor data from the robotic device, analyzes data quality metrics, and provides feedback by updating navigation policies when quality thresholds are not met. This closed-loop feedback mechanism allows the system to detect poor data quality caused by positioning errors and adjust navigation accordingly to improve future data collection.
Solution Approach 2:
The patent replaces reliance on precise mechanical positioning and orientation systems with an AI-based data quality assessment and policy update system. Instead of depending on perfect mechanical navigation, the system uses machine learning models to evaluate data quality and adapt navigation strategies, substituting mechanical precision requirements with computational intelligence.
2Ease of operation
If the robotic device follows a fixed navigation policy, then operational simplicity is maintained, but inability to adapt to environmental variations leads to poor sensing quality
Solution Approach 1:
The navigation policy is transformed from a static, fixed set of instructions to a dynamic, adaptive system that updates based on data quality analysis. The policy evolves over time by incorporating lessons from past navigation attempts and their resulting data quality outcomes, allowing the system to adapt to environmental variations while maintaining operational simplicity through automated learning.
Solution Approach 2:
The system performs self-improvement by automatically analyzing its own data quality and updating its navigation policy without external intervention. The robotic device and control system collaboratively learn from experience, with the policy updating itself based on performance feedback, enabling autonomous adaptation to environmental changes.
3Use of energy by moving object
If downstream AI pipeline resides outside the robotic device on edge or cloud servers, then energy consumption on the device is reduced, but latency and data quality evaluation delays increase
Solution Approach 1:
The patent extracts only the essential data quality evaluation functions from the downstream AI pipeline and implements them locally on the robotic device. By taking out and implementing critical quality assessment capabilities on-device, the system reduces latency for quality evaluation while maintaining energy efficiency by keeping the full AI pipeline on external servers.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Optimizing sensing capabilities of a roaming robotic device using Artificial Intelligence (Al) includes receiving data at a control system having a computer from a robotic device, the control system communicating a policy to the robotic device for choosing navigation actions for the robotic device. The received data is analyzed using the control system for determining when the received data meets a threshold upon determining quality of the data. The analysis can include generating a model based on the received data where the model includes vector representation of inputs detected by sensors at the location. In response to the received data at the control system not meeting the threshold for determining quality, the robotic device communicates with the control system to collaborate in updating the policy to choose a next action for the robotic device.