Occupancy Grid Map Change Detection for Robot Autonomy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robot architectures lack a foundation for dynamic autonomy and generic robot attributes, making it difficult to port behaviors across different platforms and rely heavily on human operators for navigation and decision-making, limiting their ability to operate independently in complex or dangerous environments.
Innovation Solution
A robot intelligence kernel with a generic robot architecture that provides a framework for dynamic autonomy, enabling seamless portability across various platforms and interaction levels, by using an occupancy grid map to detect changes and process sensor data into actionable information, allowing robots to make decisions and navigate autonomously.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a generic robot architecture is implemented to enable portability across platforms, then adaptability is improved, but device complexity increases due to the need for abstraction layers and intelligence kernel
Solution Approach 1:
The patent implements a universal robot architecture that provides a common intelligence kernel and abstraction layers capable of supporting multiple robot platforms and applications. The generic robot attributes and behaviors are designed to be platform-independent, allowing the same software architecture to control different types of robots (ground, aerial, aquatic) through standardized interfaces while maintaining the ability to adapt to specific platform requirements.
Solution Approach 2:
The architecture is divided into distinct functional layers including hardware abstraction layer, robot abstraction layer, behavior layer, and task layer. This segmentation allows each layer to be developed and modified independently, reducing the complexity burden by organizing the system into manageable modules that can be reused across different platforms without requiring complete redesign.
2Extent of automation
If robot intelligence is increased to reduce reliance on human operators, then extent of automation is improved, but device complexity increases due to decision-making capabilities
Solution Approach 1:
The intelligence kernel implements dynamic autonomy adjustment that allows the robot to adapt its level of autonomous decision-making based on environmental conditions, task requirements, and operational context. The system can dynamically switch between fully autonomous operation, semi-autonomous mode with human oversight, and teleoperated mode, managing complexity by activating only the necessary decision-making capabilities for each situation rather than maintaining all capabilities simultaneously.
Solution Approach 2:
The robot architecture includes self-service capabilities where the intelligence kernel autonomously manages its own operation, including autonomous navigation, obstacle avoidance, task execution, and even self-diagnosis. The system can make decisions about its own control without human intervention, managing its complexity through self-regulation and autonomous problem-solving rather than requiring external human management.
3Measurement precision
If occupancy change detection is implemented for autonomous navigation, then measurement precision is improved, but use of energy increases due to continuous sensor processing
Solution Approach 1:
The occupancy change detection system operates using periodic scanning rather than continuous monitoring. The robot performs sensor scans at regular intervals and compares current scans with previous occupancy maps to detect changes. This periodic operation significantly reduces energy consumption compared to continuous processing while maintaining adequate detection precision for navigation purposes by capturing environmental changes at meaningful time intervals.
Solution Approach 2:
The system processes only the portions of the environment that are relevant to current navigation tasks and occupancy changes. Rather than processing all sensor data continuously, the architecture focuses computational resources on detecting and processing only those environmental changes that affect navigation decisions, reducing overall energy usage while maintaining precision for critical detection functions.
Data Source
AI summary
A robot platform includes perceptors, locomotors, and a system controller. The system controller executes instructions for producing an occupancy grid map of an environment around the robot, scanning the environment to generate a current obstacle map relative to a current robot position, and converting the current obstacle map to a current occupancy grid map. The instructions also include processing each grid cell in the occupancy grid map. Within the processing of each grid cell, the instructions include comparing each grid cell in the occupancy grid map to a corresponding grid cell in the current occupancy grid map. For grid cells with a difference, the instructions include defining a change vector for each changed grid cell, wherein the change vector includes a direction from the robot to the changed grid cell and a range from the robot to the changed grid cell.