Autonomous Inspection Navigation Using Convex Object Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous vehicles face challenges in navigating and performing tasks in real-world environments with unknown or changing conditions, requiring improved local control and navigation systems to enhance automation and reduce operator intervention.
Innovation Solution
The development of an autonomous vehicle system that includes a multirotor helicopter configuration with a sophisticated navigation system, sensor suite, and inspection capabilities, utilizing a combination of sensors and a base station for real-time data processing and communication to refine its flight plan and ensure accurate object inspection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If the degree of automation is increased to reduce operator intervention, then productivity and safety are improved, but the system must handle unknown and changing environmental conditions which increases device complexity
Solution Approach 1:
The navigation system is divided into multiple independent modules: sensor system for environmental perception, map data system for storing known environment information, flight plan generation system for path planning, and flight control system for execution. Each module handles specific aspects of autonomous navigation, allowing the system to manage complexity through functional segmentation while achieving high automation.
Solution Approach 2:
The system performs preliminary actions by pre-generating flight plans based on stored map data before actual navigation. The sensor system continuously compares real-time environmental data with pre-stored map information, allowing the drone to anticipate and prepare for upcoming navigation decisions rather than reacting solely to immediate conditions, thus managing complexity through proactive planning.
2Ease of operation
If fully autonomous operation is implemented to reduce skilled operator requirements, then ease of operation is improved, but navigation in real-world environments with unknown objects increases device complexity
Solution Approach 1:
The system creates a digital copy of the environment through map data stored in the map data system. This virtual representation is continuously compared with real-time sensor data, allowing the autonomous vehicle to navigate by matching physical observations against the stored model. This copying approach enables fully autonomous operation by replacing operator interpretation with automated comparison between virtual and real environments.
Solution Approach 2:
The sensor system continuously provides feedback by comparing detected environmental features with corresponding data from the stored map. This feedback loop enables the autonomous vehicle to verify its location, detect deviations from the planned path, and identify unknown objects or changes in the environment, allowing ease of operation through automated environmental awareness without requiring operator intervention.
3Adaptability or versatility
If local control systems are enhanced to handle changing environmental conditions, then adaptability is improved, but device complexity increases
Solution Approach 1:
The flight plan generation system dynamically adjusts navigation plans based on real-time sensor data and changing environmental conditions. Rather than following rigid pre-programmed paths, the system continuously regenerates flight plans by comparing current sensor observations with stored map data, allowing adaptability to unknown objects and environmental changes while managing complexity through iterative re-planning rather than complex real-time control algorithms.
Data Source
AI summary
A navigation program for an autonomous vehicle, the navigation program configured to: receive an initial model of an object to be inspected by the autonomous vehicle; identify an inspection target associated with the initial model of the object; and determine an inspection location for the autonomous vehicle from which inspection target is inspectable by an inspection system of the autonomous vehicle, wherein the initial model includes one or more convex shapes representing the object.