Visual-Inertial Mapping for Low-Power Real-Time Positional Awareness
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies for providing fast, accurate, and reliable positional awareness to robots and wearable devices, such as VR/AR headsets, face challenges including high costs, power consumption issues, and limitations in recognizing locations and obstructions quickly, with existing methods like RFID/WiFi, depth sensors, and visual approaches being expensive, power-intensive, or slow.
Innovation Solution
The implementation of a visual-inertial sensor system that combines low-cost grayscale and RGB cameras with a low-power control unit, offloading computational tasks to reduce energy consumption and cost, using inertial data to correct pose estimates and process image data efficiently, enabling accurate localization and tracking in real-time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If RFID/WiFi approaches are used for positional awareness, then location recognition is enabled, but cost increases and accuracy is limited
Solution Approach 1:
The patent replaces RFID/WiFi electronic positioning systems with a visual-inertial sensing system that uses cameras and inertial measurement units to achieve positioning. This substitution enables higher accuracy through visual feature recognition and inertial data fusion while using lower-cost consumer-grade sensors compared to specialized RFID/WiFi infrastructure.
2Measurement precision
If depth sensor based approaches are used, then positional information is obtained, but cost increases and power consumption increases
Solution Approach 1:
The patent merges multiple low-power sensing modalities (grayscale camera, RGB camera, and inertial measurement unit) to achieve positioning accuracy comparable to expensive depth sensors. By combining visual data from multiple camera types with inertial data through sensor fusion, the system achieves depth perception and positional awareness without the high power consumption of active depth sensing technologies.
3Measurement precision
If marker based approaches are used, then location recognition is improved, but the operational area is limited
Solution Approach 1:
The patent extracts and removes the requirement for artificial markers from the positioning system. By using natural visual features from grayscale and RGB cameras combined with inertial navigation, the system achieves markerless positioning that works across diverse environments without requiring pre-placed markers, thereby expanding operational versatility while maintaining accuracy.
4Ease of manufacture
If visual approaches are used, then cost is reduced, but processing speed decreases leading to failure in fast motion applications
Solution Approach 1:
The patent implements preliminary action by using the inertial measurement unit to predict device motion and pre-compute expected visual feature positions before actual image capture. This allows the visual processing to start from a predicted state rather than processing raw data from scratch, significantly accelerating processing speed for fast-motion applications while using affordable visual sensors.
5Ease of manufacture
If visual approaches are used, then cost is reduced, but scale ambiguity occurs
Solution Approach 1:
The patent implements feedback by continuously fusing inertial measurement data with visual feature tracking data to resolve scale ambiguity. The inertial unit provides absolute motion reference that constrains the visual odometry solution, preventing drift and scale errors while maintaining the use of low-cost visual sensors. This closed-loop sensor fusion eliminates the scale ambiguity problem inherent in pure visual approaches.
Data Source
AI summary
The described positional awareness techniques employing visual-inertial sensory data gathering and analysis hardware with reference to specific example implementations implement improvements in the use of sensors, techniques and hardware design that can enable specific embodiments to provide positional awareness to machines with improved speed and accuracy.