Zone-Based Predictive Input System for Touch Displays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current touch and touchless display technologies face challenges in predicting user inputs accurately, particularly due to noise and latency issues, which result in prediction artifacts and reduced responsiveness, especially near the edges of the display.
Innovation Solution
A predictive input system is implemented, which divides the display into zones with dynamic look-ahead prediction values, sensitivity parameters, and algorithms based on noise levels and accuracy signatures, using interpolation and smoothing algorithms to enhance prediction accuracy and reduce latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single prediction algorithm and parameters are used across the entire display, then device complexity is reduced, but prediction accuracy deteriorates near display edges due to varying noise levels
Solution Approach 1:
The display is divided into multiple zones (e.g., center zone and edge zones) with different prediction parameters. Each zone has its own look-ahead prediction value and sensitivity settings, allowing the system to optimize prediction accuracy for each region while managing complexity through structured segmentation.
Solution Approach 2:
Different zones of the display are assigned different prediction characteristics tailored to their specific noise profiles. Edge zones use more conservative prediction parameters due to higher noise levels, while center zones can use more aggressive prediction for better responsiveness, creating local optimization without requiring a completely different system architecture.
2Productivity
If aggressive look-ahead prediction is used to improve responsiveness, then input prediction accuracy improves, but noise sensitivity increases causing prediction artifacts
Solution Approach 1:
The prediction system dynamically adjusts look-ahead values and sensitivity parameters based on the zone being accessed and the specific input context. This allows the system to be more aggressive when conditions permit (improving responsiveness) while automatically retreating to more conservative settings when noise levels would cause artifacts, creating adaptive optimization.
Solution Approach 2:
The system changes prediction parameters (look-ahead values, sensitivity thresholds) based on the operational context and zone characteristics. By adjusting these parameters dynamically rather than using fixed values, the system can optimize responsiveness when noise is low while preventing artifacts when noise increases, resolving the contradiction between aggressiveness and noise sensitivity.
3Measurement precision
If zone-based dynamic prediction parameters are implemented, then input prediction accuracy improves across different display areas, but device complexity increases
Solution Approach 1:
The display area is segmented into zones with predefined characteristics, and the system maintains a lookup table or configuration of prediction parameters for each zone. This segmentation approach improves accuracy by tailoring parameters to local conditions while controlling complexity through structured organization rather than requiring complex real-time calculations.
Solution Approach 2:
The prediction system automatically selects appropriate parameters based on the detected input zone without requiring manual configuration or complex external control. The system serves itself by using the zone identification to automatically retrieve and apply the appropriate prediction parameters, reducing the operational complexity despite the increased structural complexity of having multiple zones.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
A device comprising a display, a predictive input system coupled to the display and configured to execute a method (500) comprising the steps of: store zone data indicating zones of the display (505), wherein each zone constitutes a portion of a display area; store predictive parameter data (510), wherein each zone is assigned a subset of predictive parameter data, wherein each subset indicates at least one of a look-ahead prediction value that indicates a time period corresponding to how far in the future a predictive user input pertains, an indication of a particular prediction algorithm, or at least one value used by the particular prediction algorithm, and wherein at least two zones of the display have different values for the subsets of predictive parameter data; receive input data via the display (515); determine the zone in which the input data is received (525); and generate prediction data based on the input data, the zone data, and the subset of predictive parameter data (530).