Autonomous Vehicle Control Using Hierarchical Task Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous heavy equipment and vehicles face challenges in safely performing tasks in diverse off-road conditions due to dynamic task contexts and the complexity of emulating human decision-making, particularly in environments with varying terrain and obstacles.
Innovation Solution
An autonomous vehicle management system that uses task hierarchies, sensor data, and AI-based techniques like reinforcement learning to generate and execute plans of action, breaking tasks into unit tasks and subtasks, and selecting predictive models based on current task contexts to control vehicle operations safely.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If autonomous machines are deployed to perform tasks in diverse off-road conditions, then productivity is improved, but the difficulty of detecting and measuring task contexts and environmental conditions increases
Solution Approach 1:
The patent segments the complex autonomous control system into a hierarchical task network structure with multiple levels (parent tasks, child tasks, subtasks). Each task level handles specific aspects of operation, allowing the system to manage diverse off-road conditions by breaking down complex environmental detection into manageable task-specific components rather than requiring a single monolithic detection system.
Solution Approach 2:
The system performs preliminary actions by pre-defining task hierarchies and selecting appropriate predictive models before executing tasks. The controller system identifies required predictive models based on task context in advance, preparing the appropriate detection and measurement capabilities before encountering specific environmental conditions, thus improving productivity without increasing real-time detection complexity.
2Adaptability or versatility
If human decision-making processes are emulated in autonomous systems, then adaptability to changing environments is improved, but device complexity increases
Solution Approach 1:
The patent replaces complex mechanical decision-making emulation with a structured task hierarchy system guided by predictive models. Instead of attempting to replicate human cognitive processes, the system uses predetermined task networks combined with context-aware model selection to achieve adaptability. This substitution reduces device complexity by replacing sophisticated software-based decision emulation with a more manageable hierarchical control structure.
Solution Approach 2:
The system achieves adaptability through parameter changes in the form of selecting different predictive models based on task context. Rather than implementing a universally complex decision-making system, the controller dynamically adjusts which predictive model is active based on environmental parameters and task requirements. This allows the system to adapt to changing environments while maintaining relatively simple control logic at any given moment.
3Reliability
If multiple predictive models are used to handle different task contexts, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent implements a dynamic model selection mechanism where the controller system identifies and activates the appropriate predictive model based on the current task context. The hierarchical task network structure allows the system to dynamically adjust which models are relevant for each task level, ensuring reliable operation across diverse conditions while managing model complexity through context-based activation rather than maintaining all models simultaneously active.
Solution Approach 2:
The hierarchical task network structure serves as an intermediary between multiple predictive models and the control system. Rather than having the controller directly manage numerous models, the task hierarchy acts as an intermediate layer that selects and coordinates appropriate models based on task requirements. This intermediary structure improves reliability by ensuring the right model is used for each context while reducing the apparent complexity for the control system.
Data Source
AI summary
The present disclosure relates techniques for autonomously controlling heavy equipment and vehicles using task hierarchies. Particularly, aspects of the present disclosure are directed to obtaining a task to be performed by an autonomous vehicle, determining subtasks to be performed to perform the task, obtaining sensor data providing a representation of operation of the autonomous vehicle in a worksite environment and situational context of the worksite environment, determining a task context for a subtask based on the sensor data, identifying a predictive model from a library of predictive models based on the task context, estimating, by the predictive model, a set of output data based on sensor data, and controlling operations of the autonomous vehicle in the worksite environment to perform the subtask using a set of input data derived from the sensor data and the set of output data.