Direction Determination Using Multi-State Sensor Inputs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining the direction of a moving object, such as a vehicle, using truth tables are limited by their inability to handle unknown sequences, and neural networks are expensive and complex to reconfigure.
Innovation Solution
A computer-implemented method that monitors state changes across a sensor with multiple inputs, comparing the magnitude of changes to determine the direction of a moving object by incrementing or decrementing a direction variable based on the order of state transitions, allowing for accurate direction assignment without requiring all inputs to be monitored simultaneously.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If truth tables are used to determine direction, then the system is simple to implement, but it cannot handle unknown sequences and lacks accuracy
Solution Approach 1:
The system dynamically adapts by learning the optimal truth table configuration through training data rather than using a static, pre-defined truth table. The truth table is updated based on learned patterns from training examples, allowing the system to handle unknown sequences while maintaining simplicity.
Solution Approach 2:
The system changes the parameters of the truth table based on learned characteristics from training data. By adjusting the truth table configuration dynamically according to learned patterns, the system improves accuracy for unknown sequences while retaining the simplicity of truth table-based operation.
2Reliability
If neural networks are used to determine direction, then accuracy is high, but the system is expensive and complex to reconfigure
Solution Approach 1:
The invention extracts the essential learning capability from complex neural networks and implements it through a simplified truth table structure. By taking out only the necessary adaptive function and embedding it in a simple truth table, the system achieves high accuracy without the complexity and cost of full neural networks.
Solution Approach 2:
The system uses a simple, inexpensive truth table structure that can be easily reconfigured through software updates rather than requiring expensive, complex neural network hardware. The truth table acts as a lightweight, easily replaceable component that provides neural-network-like accuracy at fraction of the cost.
3Loss of information
If all inputs are monitored simultaneously, then complete information is captured, but the system complexity increases
Solution Approach 1:
The system monitors inputs selectively rather than all inputs simultaneously. By using the learned truth table to identify which input combinations are relevant, the system captures sufficient information to determine direction without the complexity of monitoring all possible inputs at all times.
Data Source
AI summary
A computer-implemented method for determining the direction of a moving object across a sensor having a plurality of inputs is disclosed. The invention determines the direction of a moving object, such as a vehicle on a roadway, based on inputs provided by sensors along the object's path. The methods involve monitoring state changes of the inputs as the object passes the sensor and comparing the magnitude of state changes in the order in which a forward-moving object would cause inputs to switch to a particular state. For each state change comparison, a direction variable is adjusted to indicate forward or reverse movement. After passage of the object, a direction of movement is assigned to the moving object on the basis of the final value of the direction variable. The invention provides a high degree of accuracy, is simple to reconfigure, and more economical than other methods.