Finite-State Automata for Optical Target Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automation methods using finite-state automata with optical sensors are inflexible, costly, and complex, requiring separate control systems, leading to high latency and management costs, and are not easily adaptable to different scenarios.
Innovation Solution
A method employing imaging devices and neural network-based algorithms to capture and process environmental information, enabling real-time control of apparatuses by recognizing targets and their states, and using finite-state automata to manage and control apparatuses based on this information, reducing the need for multiple optical sensors and specialized components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If finite-state automata with optical sensors are used for automation, then automation capability is achieved, but device complexity and costs increase due to requiring one optical sensor per automaton
Solution Approach 1:
The patent implements a single optical sensor that serves multiple finite-state automata simultaneously. Instead of dedicating one optical sensor to each automaton, the system allows one sensor to be shared across multiple automata, reducing the total number of sensors required while maintaining full automation capability for all monitored conditions.
Solution Approach 2:
The patent merges the functions of multiple optical sensors into a single sensor by combining the signal processing paths of multiple automata. The control system integrates inputs from a single sensor to trigger multiple automata based on different signal characteristics, effectively combining what would have been separate sensing functions into one unified sensor system.
2Extent of automation
If separate control and analysis systems are used for high complexity automation, then automation functionality is achieved, but response time increases due to additional latency
Solution Approach 1:
The patent merges the control and analysis functions into a single integrated system. The control unit directly processes signals from the optical sensor and triggers automata without requiring separate analysis systems, eliminating the latency introduced by inter-system communication and data transfer between independent control and analysis components.
3Measurement precision
If specialized components are used for specific scenarios, then automation precision is improved, but adaptability to other scenarios decreases
Solution Approach 1:
The patent implements a universal control system that can manage multiple finite-state automata with a single optical sensor. The system is designed to be scenario-agnostic, allowing the same hardware configuration to be adapted to different automation scenarios by simply reconfiguring the automata logic and signal processing parameters, thereby maintaining precision across various applications without requiring scenario-specific specialized components.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method (1) of managing one or more apparatuses (100) comprising the steps of: a) cyclically capturing, by an imaging device (200), a corresponding image stream associated with a field of view (11) as viewed by the imaging device (200), the captured image stream comprising a plurality of temporally successive frames; b) sending the image stream to a data processing unit (210) associated with the imaging device (200); c) recognizing one or more targets (12) in the field of view (11) that has been captured in the image stream and a target state (St) for each recognized target (12) representative of the target (12); d) associating each recognized target (12) with a label (E) and a finite-state automaton (A) as a function of the corresponding label (E), each finite-state automaton (A) comprising two or more automaton states (SA), automaton state transition signals (SpA) for transition from one automaton state to another and automaton holding signals (SmA) for holding the automaton state (SA) and action signals (Sc) for managing, enabling and/or disabling one or more apparatuses (100); e) initializing each finite-state automaton (A) to an initial automaton state (SiA) selected from the states of the finite-state automaton (SA) as a function of the target state (St); f) generating an assessment signal (Sv) for each recognized target (12) as a function of the target state (St) and as a function of temporally successive frames; g) triggering each initialized finite-state automaton (A) by: g1) transitioning from the initial automaton state (SiA) to an altered automaton state (SfA) or from an altered automaton state (SfA) to another altered automaton state (SfA); or g2) holding the initial automaton state (SiA) or the altered automaton state (SfA); h) sending an action signal (Sc) to one or more apparatuses (200) for each finite-state automaton (A); i) managing the one or more apparatuses (200) according to the action signal that has been sent (Sc).