Autonomy Framework With Mission Graphs for Cross-Platform Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robotics and autonomy technologies are platform-specific, lacking scalability and reusability, and fail to enable effective collaboration and customization of missions across various unmanned platforms, leading to vendor lock-in and duplicative development efforts.
Innovation Solution
The Cross Platform Autonomy Stack (CPAS) framework, comprising a Mission Behavior Manager (MBM) and Autonomous Action Controller (AAC) modules, translates mission scripts into actionable commands and interfaces with vendor-specific peripherals, sensors, and payloads, allowing deployment on multiple autonomous vehicle platforms through a connected graph data structure and plug-in library validation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If autonomy frameworks are platform-specific, then they can be optimized for specific hardware, but they lack scalability and reusability across different unmanned platforms
Solution Approach 1:
The patent implements a universal autonomy framework that can operate across multiple unmanned platform types (aerial, ground, maritime) through a common architecture. The framework uses standardized interfaces and abstraction layers that allow the same autonomy algorithms to function on different hardware platforms without requiring platform-specific implementations, thereby achieving multi-functionality and broad adaptability.
Solution Approach 2:
The autonomy framework is segmented into modular components including perception modules, navigation modules, mission management modules, and communication modules. Each module can be independently configured and deployed based on platform requirements. This segmentation allows the framework to maintain complexity manageability while achieving versatility across platforms through selective module assembly.
2Productivity
If autonomy algorithms are customized for each platform, then they can achieve optimal performance, but duplicative development efforts increase and collaboration among network participants decreases
Solution Approach 1:
The patent implements a centralized mission library that stores standardized mission profiles and algorithms developed once can be copied and deployed across multiple platforms. This eliminates duplicative development efforts while maintaining reliability through rigorous validation of copied missions. The framework includes mechanisms for mission verification and adaptation that ensure copied missions execute reliably on target platforms.
Solution Approach 2:
The framework incorporates feedback mechanisms where mission execution data from different platforms is collected and analyzed to improve algorithms. This continuous feedback loop allows the system to maintain high reliability across platforms without requiring custom development for each one, as lessons learned from one platform's operations improve the general algorithm suite available to all platforms.
3Ease of operation
If a standardized command interface is implemented, then ease of operation increases, but the ability to handle vendor-specific peripherals and sensors may be compromised
Solution Approach 1:
The patent introduces an intermediary layer consisting of standardized adapters and drivers that sit between the universal command interface and vendor-specific hardware. These intermediaries translate standardized commands into platform-specific operations, allowing users to operate the system through a simple standardized interface while the intermediary layer handles the complexity of vendor-specific peripherals and sensors transparently.
Data Source
AI summary
Embodiments relate to an autonomous action deployment system having one or more processors. The system includes a mission behavior manager (MBM) module and an autonomous action controller (AAC) module. A processor is configured to interface with one or more autonomous vehicle platforms. The MBM module is configured to receive a mission script that includes a text file, parse the text file into actions, create a data structure where each action exists as a node in a connected graph, validate that each node corresponds with a plug-in stored in a plug-in library specific to an autonomous vehicle platform, and generate an action request command. The AAC module is configured to receive the action request command and execute steps to accomplish the action request command for the autonomous vehicle platform via a corresponding driver.