AR Sensor Guidance for Flow Field Measurement Uncertainty
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In engineering applications, particularly in aerodynamics, it is challenging to accurately forecast and measure complex flow fields due to substantial uncertainties in boundary conditions, leading to uncertain results in physics simulations.
Innovation Solution
A system combining pose determination of handheld sensors with augmented reality and sensor location optimization software, using Active Learning algorithms to provide real-time visualization for optimal data gathering, allowing operators to move sensors to optimal locations for improved measurement results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If handheld sensors are manually moved to different locations for data gathering, then measurement coverage is improved, but measurement precision and data quality are degraded due to lack of optimal location selection
Solution Approach 1:
The system continuously receives sensor readings, processes them through Active Learning algorithms to determine uncertainty, and provides real-time feedback via AR visualization showing optimal next measurement locations. This closed-loop feedback enables operators to systematically improve data quality by navigating to high-uncertainty areas identified by the algorithm.
Solution Approach 2:
The patent replaces manual expert judgment and mechanical sensor positioning with an automated computational system using Active Learning algorithms and AR visualization. The system automatically determines optimal measurement locations and guides operators, substituting human expertise with algorithmic decision-making.
2Productivity
If physics simulations are run to predict flow fields, then design optimization is improved, but reliability is degraded due to substantial uncertainties in boundary conditions
Solution Approach 1:
The system enables the measurement process to self-optimize by automatically identifying high-uncertainty regions and guiding sensor placement to those areas. The Active Learning algorithm autonomously determines where measurements are most needed, allowing the system to self-correct and improve simulation reliability without external intervention.
Solution Approach 2:
The system dynamically changes measurement parameters by adapting sensor placement locations based on real-time uncertainty analysis. Instead of fixed measurement grids, the system continuously adjusts measurement locations to target areas of highest uncertainty, thereby improving the reliability of flow field predictions.
3Ease of operation
If environmental sampling is performed without optimization, then ease of operation is improved, but productivity is degraded due to inefficient data collection
Solution Approach 1:
The AR visualization acts as an intermediary between the complex Active Learning algorithm and the operator. It translates algorithmic uncertainty calculations into intuitive visual guidance, maintaining ease of operation while dramatically improving productivity through optimized measurement paths.
Data Source
AI summary
Systems and methods for real-time environmental sensor data gathering is enhanced using augmented reality, with a virtual target object being presented to the user of the sensor device that guides the user where to move the sensor device next. A combination of pose data for the sensor and data modeling of the sensor data allows for users with minimal training to make optimized environmental readings.