Differentiated Multi-Agent Navigation for GPS-Free Pose Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation systems, particularly GPS-based systems, are ineffective in GPS-denied environments, where agents cannot rely on satellite signals for navigation.
Innovation Solution
The development of computer-implemented methods, media, and systems for multi-agent navigation that utilize relative distance data, anomaly sensor data, and pre-surveyed map data to determine global pose and assign tasks based on specialized operational capabilities, without relying on external systems like GPS.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS-based navigation is used, then navigation accuracy is improved, but the system becomes ineffective in GPS-denied environments
Solution Approach 1:
The patent introduces anomaly sensor subsystems as intermediary devices that detect environmental anomalies (magnetic, gravitational, electromagnetic) to serve as mediators for navigation in GPS-denied environments. These sensors act as intermediaries between the agents and the environment, providing alternative reference points for pose estimation when GPS signals are unavailable.
Solution Approach 2:
The patent replaces the GPS satellite-based electromagnetic signal system with a local anomaly detection system. Instead of relying on external satellite signals, the system substitutes a local navigation approach using anomaly sensors to detect and map environmental features, thereby substituting the external GPS mechanism with an internal anomaly-based navigation mechanism.
2Reliability
If all agents are equipped with anomaly sensor subsystems, then navigation capability in GPS-denied environments is improved, but device complexity and cost increase
Solution Approach 1:
The patent segments the agent population into different subsets: a first subset equipped with anomaly sensor subsystems for navigation, and a second subset without such sensors. This segmentation allows the system to distribute the navigation function across only the necessary agents, reducing overall system complexity while maintaining navigation capability through cooperative multi-agent operation.
Solution Approach 2:
The patent creates universality by enabling agents without anomaly sensors to still participate in navigation tasks through cooperative multi-agent navigation. The second subset of agents, though lacking anomaly sensors, can still navigate by receiving and processing navigation commands from the first subset, making the navigation system universal across all agent types.
3Measurement precision
If cooperative multi-agent navigation is implemented, then absolute pose estimation accuracy is improved, but communication and coordination requirements increase
Solution Approach 1:
The patent merges the navigation computations across multiple agents by combining relative distance measurements from multiple agent pairs with anomaly data. The navigation system integrates information from multiple sources (relative distance data from different agent pairs, anomaly sensor readings) to collectively determine absolute pose, achieving higher accuracy through merged information than any single agent could achieve alone.
Data Source
AI summary
Example computer-implemented methods and systems for anomaly-sensing based multi-agent navigation are disclosed. One example computer-implemented method includes: receiving relative distance data specifying distance between at least one pair of agents of a plurality of agents, each of a first subset of the plurality of agents having an anomaly sensor subsystem; determining a set of relative pose vectors based at least in part on the relative distance data; receiving anomaly data from at least one anomaly sensor subsystem of one of the plurality of agents; obtaining pre-surveyed map data; determining global pose data of the plurality of agents based on the relative distance data and based on comparing the anomaly data to the pre-surveyed map data; and assigning a task to at least one of the plurality of agents based at least in part on a specialized operational capability of the at least one of the plurality of agents.