Robot Task Graph Execution for Contingency-Adaptive Missions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomy systems for robots are limited in extensibility as they are typically configured to address only one narrow mission set, lacking the ability to adapt to unexpected contingency tasks, which can lead to inefficiencies and safety concerns in complex environments.
Innovation Solution
A method and system that allows robots to execute missions by monitoring their environment and transitioning between tasks based on detected contingency levels, using a task graph and library to dynamically adjust maneuvers and manage contingency events, enabling adaptive mission execution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If existing autonomy systems are configured to address only one narrow mission set, then the system design can be simplified and optimized for specific tasks, but the system lacks extensibility and cannot adapt to unexpected contingency tasks
Solution Approach 1:
The autonomy system is segmented into modular components including a task library with independent task definitions, a task graph structure that organizes tasks hierarchically, and a mission manager that coordinates between components. This modular segmentation enables the system to extend functionality by adding new tasks without redesigning the entire system, resolving the contradiction between adaptability and complexity.
Solution Approach 2:
The task library employs universal task definitions and standardized interfaces that can represent multiple types of operations (navigation, manipulation, observation) through a common framework. This universality allows the same system architecture to handle diverse mission types and contingency scenarios, achieving extensibility without proportionally increasing system complexity.
2Reliability
If the robot immediately executes contingency tasks when detected, then the robot can respond quickly to unexpected situations, but this may waste time if the contingency task does not actually impact the mission
Solution Approach 1:
The system performs preliminary analysis of detected contingency events by evaluating their potential impact on mission objectives before executing contingency tasks. The mission manager assesses whether the contingency actually affects mission completion, and only then triggers the corresponding contingency task. This preliminary evaluation prevents unnecessary task execution and time wastage while maintaining reliable mission completion.
3Productivity
If the robot defers action on potential contingency tasks, then the robot avoids wasting time on unnecessary tasks, but this may incur penalties in terms of wasted time if the contingency task becomes a bottleneck
Solution Approach 1:
The system performs preliminary assessment of contingency potential by analyzing the detected event's characteristics, the current mission state, and the task graph structure. This preliminary action identifies which contingency tasks are likely to become bottlenecks and should be executed promptly, versus those that can be safely deferred. This resolves the contradiction by using intelligent preliminary evaluation to optimize mission efficiency while minimizing idle time.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
A method and corresponding apparatus and computer-readable storage medium are provided for causing one or more robots to execute a mission. The method includes identifying the mission including a nominal sequence of selected tasks that are executable to cause the one or more robots to execute maneuvers to achieve a mission objective. The method includes determining a task graph in which the mission is modeled. The task graph is expressed as a directed graph and includes selected task nodes representing the selected tasks that are connected by edges representing transitions between the selected tasks. The method also includes causing the one or more robots to execute the mission using the task graph and a task library of tasks including a selected task executable to cause the one or more robots to execute a maneuver.