System and method for assessing fatigue of pilot of aircraft

Abstract

The invention relates to a system and method for assessing fatigue of a pilot of an aircraft. A system (100) and method includes an artificial intelligence (AI) control unit (102) configured to receive a schedule for a pilot of an aircraft (106), receive survey answer data (128) from the pilot, analyze the schedule and the survey answer data (128) using one or more machine learning models (124), and transmit the analyzed schedule and survey answer data (128) to the aircraft (106). And evaluating a readiness level of the pilot based on the schedule and survey answer data (128). The aircraft (106) is operated according to the readiness state level evaluated by the AI control unit (102).

Description

Technical Field

[0001] Examples of the present disclosure generally relate to systems and methods for assessing the readiness of a pilot of an aircraft. Background Art

[0002] Aircraft transport passengers and cargo between different locations. Many aircraft depart from and arrive at a typical airport every day.

[0003] Pilots operate aircraft according to a schedule. That is, the flight schedule includes the dates and times when the pilot operates the aircraft between different airports. It can be understood that if too many flights are planned within a short period of time, the pilot may become fatigued.

[0004] The pilot's readiness for flight (such as rest, fatigue level) is related to flight operations. For example, high fatigue may lead to a decline in performance, poor decision-making, and an increased risk of accidents. There are several factors that contribute to the pilot's readiness (including time of day, duration of the flight, multiple time zones crossed, etc.). In addition, individual differences such as sleep patterns, sleep disorders, and stress levels may also contribute to pilot fatigue.

[0005] The current self-assessment process for pilot fatigue is crude and typically relies on asking the pilot a question about the perceived level of readiness. Traditional fatigue risk management tools, while effective in some cases, may not fully meet the unique needs of pilots. Summary of the Invention

[0006] There is a need for a system and method for accurately assessing a pilot's readiness. In addition, there is a need for customized methods for predicting and guiding readiness.

[0007] In view of these needs, certain examples of the present disclosure provide a system that includes an artificial intelligence (AI) control unit configured to receive a schedule for a pilot of an aircraft, receive survey answer data from the pilot, analyze the schedule and the survey answer data using one or more machine learning models, and evaluate a level of the pilot's readiness based on the schedule and the survey answer data. The aircraft is operated according to the level of readiness evaluated by the AI control unit.

[0008] In at least one example, the system further includes a user interface in communication with the AI control unit. The user interface includes a display in communication with an input device. The AI control unit displays one or more surveys on the display. The pilot inputs answers to questions in the one or more surveys via the input device. The survey answer data includes the answers. The user interface may be on the aircraft.

[0009] The system may include a survey database that communicates with an AI control unit. The survey database stores survey data including one or more surveys.

[0010] The system may include a schedule database that communicates with the AI control unit. The schedule database stores schedule data including a schedule.

[0011] The system may include a model database that communicates with the AI control unit. The model database stores one or more machine learning models.

[0012] In at least one example, the one or more machine learning models may include multiple machine learning models. The AI control unit may also be configured to select the result of one of the multiple machine learning models based on an evaluated reliability.

[0013] In at least one example, the one or more machine learning models include a voting classifier model.

[0014] The AI control unit may also be configured to automatically operate one or more aspects of an aircraft based on a readiness level evaluated by the AI control unit.

[0015] Certain examples of the present disclosure provide a method that includes receiving, by an artificial intelligence (AI) control unit, a schedule for a pilot of an aircraft; receiving, by the AI control unit, survey answer data from the pilot; analyzing, by the AI control unit, the schedule and the survey answer data using one or more machine learning models; and evaluating, by the AI control unit, a readiness level of the pilot based on the use, wherein the aircraft is operated according to the readiness level evaluated by the AI control unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A block diagram of a system according to an example of the present disclosure is shown.

[0017] Figure 2 A flowchart of a method according to an example of the present disclosure is shown.

[0018] Figure 3 A front view of a display of a user interface according to an example of the present disclosure is shown.

[0019] Figure 4 A front view of a display of a user interface according to an example of the present disclosure is shown.

[0020] Figure 5 A schematic block diagram of a control unit according to an example of the present disclosure is shown.

[0021] Figure 6A perspective front view of an aircraft according to an example of the present disclosure is shown. Detailed Description

[0022] The above summary and the following detailed description of certain examples will be better understood when read in conjunction with the accompanying drawings. As used herein, an element or step recited in the singular and preceded by the word "a" or "an" should not necessarily be construed as excluding a plurality of elements or steps. Further, a reference to "one example" is not intended to be construed as excluding the existence of additional examples that also incorporate the recited features. Additionally, unless expressly stated to the contrary, an example that includes one or more elements having a particular condition may include additional elements that do not have that condition.

[0023] Examples of the present disclosure provide systems and methods for evaluating a pilot's readiness. The systems and methods provide a non-invasive solution for predicting a pilot's readiness using machine learning (e.g., via a voting classifier and engineered features). The systems and methods include a control unit that receives historical pilot schedules and readiness self-assessment data. In at least one example, the control unit employs various data analysis techniques (such as correlation analysis) and data exploration to identify patterns indicative of a pilot's readiness. In addition to these, dimensionality reduction, feature scaling, and feature selection techniques can also be used to optimize the performance of a particular model.

[0024] Figure 1 A block diagram of a system 100 according to an example of the present disclosure is shown. System 100 includes an artificial intelligence (AI) control unit 102 that communicates with one or more user interfaces 104, which allow an individual such as a pilot to input information. As shown, the user interface 104 can be on an aircraft 106 that travels between one or more departure airports and one or more arrival airports. Optionally, the user interface 104 can be remote from the aircraft 106, such as at a land-based location (e.g., an airport lounge, the pilot's home, etc.).

[0025] Each aircraft 106 includes control means 108 configured to allow an operator (e.g., a pilot) to control the operation of the aircraft 106. For example, the control means 108 includes one or more of a control handle, a yoke, a joystick, control surface control means, an accelerator, a decelerator, etc.

[0026] In at least one example, the user interface 104 can be in the cockpit or cabin of the aircraft 106. The user interface 104 includes a display 110 and an input device 112. The display 110 can be a monitor, screen, television, touch screen, etc. The input device 112 can include a keyboard, mouse, stylus, touch screen interface (i.e., the input device 112 can be integrated with the display 110), etc. The user interface 104 can be a computer workstation or a part thereof. For example, the user interface 104 can be a part of a flight computer in the cockpit or cabin of the aircraft 106. As another example, the user interface 104 can be a handheld device, such as a smartphone, tablet computer, etc.

[0027] The AI control unit 102 can be separate and distinct from the aircraft 106. For example, the AI control unit 102 can be in a central monitoring location (such as at an airport) or a location away from the airport.

[0028] The AI control unit 102 also communicates with a schedule database 114 (e.g., via one or more wired or wireless connections). The schedule database 114 stores schedule data 116 for the crew of the aircraft 106, where the crew includes the pilots of the aircraft 106. The schedule data 116 includes the flight schedule of the pilots within a predetermined time. For example, the schedule data 116 includes the pilots' schedules for recent flights (such as the previous week, previous month, or previous year or past flights) and future flights (such as the next week, next month, or next year of future flights).

[0029] The AI control unit 102 also communicates with a survey database 118 (e.g., via one or more wired or wireless connections). The survey database 118 stores survey data 120, which includes surveys with various questions related to the pilots' readiness states (e.g., regarding alert levels, rest, fatigue levels, etc.).

[0030] The AI control unit 102 also communicates with a model database 122 (e.g., via one or more wired or wireless connections). The model database 122 stores one or more machine learning models 124.

[0031] The AI control unit 102 can be co-located with one or more of the schedule database 114, survey database 118, and / or model database 122. As another example, the AI control unit 102 can be remotely located from the schedule database 114, survey database 118, and / or model database 122. In at least one other example, the AI control unit 102 can be on the aircraft 106.

[0032] In at least one example, the AI control unit 102 may also communicate with the control device 108 of the aircraft 106 and be configured to automatically operate one or more control devices 108 as described herein. Optionally, the AI control unit 102 may not communicate with the control device 108 and may not be configured to automatically operate the aircraft 106.

[0033] In operation, the AI control unit 102 outputs a survey to the user interface 104. The survey is stored in the survey data 120 and includes a number of questions for the pilot of the aircraft 106 to answer. The AI control unit 102 presents the survey including the questions on the display 110. The pilot uses the input device 112 to answer the questions in the survey. The answers to the survey questions are output to the AI control unit 102 as survey answer data 128. The survey answer data 128 includes the answers from the pilot related to the questions of the survey.

[0034] The AI control unit 102 also receives the schedule of the pilot who answered the survey questions. The schedule is within the schedule data 116 of the schedule database 114.

[0035] The AI control unit 102 analyzes the survey answer data 128 and the pilot's schedule to evaluate the pilot's readiness level. In at least one example, the AI control unit 102 analyzes the survey answer data 128 and the pilot's schedule based on one or more machine learning models 124 stored in the model database 122. Examples of the machine learning models 124 include logistic regression, Gaussian naive Bayes (NB), random forest, support vector (SV) classifier, decision tree, K-nearest neighbor (KNN), multi-layer perceptron (MLP) classifier, adaptive boosting (ADABoost) classifier, gradient boosting (GB) classifier, and / or extra tree classifier. Other types of machine learning models may be used.

[0036] In at least one example, the AI control unit 102 analyzes the survey answer data 128 and the schedule based on a single machine learning model 124 to evaluate the pilot's readiness level. In at least one other example, the AI control unit 102 analyzes the survey answer data 128 and the schedule based on multiple machine learning models 124 (e.g., two or more of the above examples). In this example, the AI control unit 102 may evaluate the data based on multiple machine learning models 124 and select the result of one of the machine learning models 124 with an evaluated, predetermined, and / or desired reliability (e.g., based on accuracy score, precision score, and / or recall score) to evaluate the pilot's readiness level.

[0037] After the AI control unit 102 uses one or more machine learning models 124 to evaluate the pilot's readiness level by analyzing the survey answer data 128 and the pilot's schedule, the AI control unit 104 can display the evaluated readiness level on the display 110. The evaluated readiness level can be displayed as a probability (e.g., percentage). The evaluated readiness level can be displayed as a color-coded indication. The evaluated readiness level can be displayed as a pre-determined readiness level (such as low, medium, or high) along with a recommendation on whether and / or how the pilot should operate the aircraft 106 (such as via an autopilot assist).

[0038] The aircraft 106 is operated based on the pilot's evaluated readiness level. For example, if the AI control unit 102 evaluates a low fatigue level for the pilot, the pilot can operate the aircraft 106 in a normal manner. As another example, if the AI control unit 102 evaluates a high fatigue level for the pilot, the pilot can be prevented from operating the aircraft 106.

[0039] In at least one example, one or more operators of the aircraft can use the readiness levels of one or more pilots to operate one or more aspects of the aircraft. The aircraft 106 can be operated according to the pilot's readiness level. For example, a pilot with a high level of readiness (e.g., little or no fatigue) can operate the aircraft 106 with few flight restrictions. As the readiness level decreases, the automated operation of the control device 108 (e.g., via an autopilot system and / or via the AI control unit 102) can increase. Additionally, the pilot's readiness level can be output to a dispatcher, flight operations personnel, etc., who can then schedule one or more pilots for a flight. In at least one example, the AI control unit 102 can automatically provide a schedule based on the pilot's readiness level (e.g., the AI control unit 102 can reduce the flights of pilots showing a higher level of fatigue).

[0040] In at least one example, the AI control unit 102 may automatically operate one or more aspects of the aircraft 106 (e.g., via one or more control devices 108) based on the pilot's evaluated readiness level. For example, if the evaluated readiness level is unusually high, the AI control unit 102 may prevent the pilot from operating the control device 108. As another example, the AI control unit 102 may automatically operate one or more control devices 108 (e.g., automatically operate the autopilot feature / function during the cruise of a staged flight) based on the pilot's evaluated readiness level. As another example, the AI control unit 102 may automatically operate an automatic flight control assist device based on the pilot's evaluated readiness level. Optionally, the AI control unit 102 may not be configured to automatically operate the aircraft 106.

[0041] As described herein, the system 100 includes an AI control unit 102 that is configured to receive one or more schedules for one or more pilots of one or more aircraft 106. The AI control unit 102 is also configured to receive survey answer data 128 from the pilots and use one or more machine learning models 124 to analyze the schedules and the survey answer data 128. The AI control unit 102 is also configured to evaluate one or more readiness levels of the pilots based on the schedules and the survey answer data 128. The aircraft 106 is operated according to the one or more readiness levels evaluated by the AI control unit 102.

[0042] The user interface 104 communicates with the AI control unit 102. The user interface 104 includes a display 110 that communicates with an input device 112. The AI control unit 102 displays one or more surveys on the display 110. The pilot inputs answers to the questions in the survey via the input device 112. The survey answer data 128 includes the answers.

[0043] Figure 2 A flowchart of a method according to an example of the present disclosure is shown. Referring Figure 1 and Figure 2 , at 200, the AI control unit 102 receives a schedule for a pilot of the aircraft 106 (including in the schedule data 116). At 202, the AI control unit receives the survey answer data 128 from the pilot. At 204, the AI control unit uses one or more machine learning models 124 to analyze the schedule and the survey answer data 128. At 206, the AI control unit 102 evaluates the pilot's readiness level based on the analysis of the schedule and the survey answer data 128 in 204.

[0044] In at least one example, the AI control unit 102 uses multiple machine learning models 124 to analyze the schedule and survey answer data 128. Additionally, in at least one example, the AI control unit 102 uses a voting classifier method trained on historical pilot schedules and readiness self-assessment data to predict the pilot's readiness level. In doing so, the AI control unit 102 overcomes the limitations of existing methods by providing a more tailored solution that takes into account the specific needs of the pilot while also eliminating, minimizing, or otherwise reducing the need for invasive physiological data collection.

[0045] As described above, the AI control unit 102 receives the schedule for the pilot. The data in the schedule includes information about flight duration, flight interval time, rest periods, and other relevant factors such as crew size, mission complexity, weather conditions during the flight, etc. For each flight of the pilot, the data about the schedule can be received and stored in the schedule database 114. The information in the pilot's schedule can include flight date, pilot identification information, departure airport, arrival airport, flight type, co-pilot information, other crew member information, crew size, takeoff time, landing time, aircraft model, etc. In at least one example, the survey data 120 includes one or more electronic surveys that provide a series of questions developed by subject matter experts that identify risks based on the Federal Aviation Administration (FAA) Pilot, Aircraft, Environment, and External Pressure (PAVE) checklist.

[0046] Figure 3 A front view of a display 110 of a user interface 104 according to an example of the present disclosure is shown. Figure 4 A front view of a display 110 of the user interface 104 is shown. Reference Figures 1-4, the AI control unit 102 displays an electronic survey on the display 110. The survey data 120 in the survey database 118 includes the electronic survey. In at least one example, the survey displayed on the display 110 includes a topic 300, questions 302 about the topic 300, and answer areas 304 for the questions 302. For example, the survey includes the topic 300 of the crew, and questions about the pilots regarding the recency, proficiency, and crew combination of the flight. The pilots use the input device 112 to answer each question 302 with a selected answer from the answer area 304. As another example, the survey includes the topic 300 of the mission, and questions about mission complexity, mission changes, and alterations. Additional topics, questions, and answer areas may be presented. Other examples include whether the crew is in training, mechanical problems (such as regarding seats, flight controls, computers, etc.), configuration changes, delays, crew pairing, duty, consecutive work days, runway conditions, weather conditions, collision risks, intangibles, etc. The above are merely examples of the questions and are non - restrictive. The pilots answer each question presented on the display 110, and the AI control unit 102 receives the answers in the survey answer data 128.

[0047] In at least one example, the AI control unit 102 combines the pilot's schedule and the survey answer data 128 to identify characteristics specifically related to the pilot's readiness state. In this way, the AI control unit 102 is configured to associate the historical schedule with the readiness state level for each flight of the pilot.

[0048] In at least one example, the AI control unit 102 uses domain knowledge to identify factors that affect the pilot's readiness state, such as flight duration, flight interval time, rest periods, and other risk factors. These variables are incorporated into the machine learning model 124 and capture potential patterns in the pilot's readiness state. Non - restrictive examples of factors that affect the pilot's readiness state include the number of crew members, the date and time of the flight, the time since the last flight, the rolling average of flight hours for the last 5 - 10 flights, the number of flights before the current flight, the number of flights in a day, a week, a month, and / or a year, the trend of flight hours for the previous several flights (such as 5 or 10 flights), the type of flight scheduled, the airport scheduled, the time to the next scheduled flight, etc. In at least one example, the AI control unit 102 also uses correlation analysis to identify the relationships between different characteristics and their contributions to the pilot's readiness state.

[0049] In at least one example, the AI control unit 102 employs data exploration techniques such as visualization and descriptive statistics to gain insights into the data and identify any patterns or trends that may be useful for the machine learning model 124. This process helps to improve the feature set and guide additional feature engineering steps. For example, the AI control unit 102 can determine the percentage of features that contribute to the overall risk. For example, the AI control unit 102 can determine that task complexity, planned task test risk, and intangible considerations contribute the most to the overall risk. As another example, the AI control unit 102 can determine that the percentage of medium / high-risk flights is higher during early morning and late-night flights than at other times. As another example, the AI control unit 102 can determine that the percentage of medium / high-risk flights is slightly higher on Mondays, Tuesdays, and Thursdays compared to other weekdays. As another example, the AI control unit 102 can determine that the percentage of medium / high-risk flights is higher at a specific airport.

[0050] In at least one example, the AI control unit 102 employs dimensionality reduction techniques to reduce the number of features in the dataset while retaining the desired or other important information. Dimensionality reduction techniques help to reduce noise and improve model performance.

[0051] In at least one example, the AI control unit 102 employs principal component analysis (PCA), which is a linear transformation technique that projects the original data onto a lower-dimensional space. PCA aims to identify the directions (or principal components) that capture the maximum variance in the data. In at least one example, the AI control unit 102 performs PCA to determine whether the survey is sufficient for obtaining answers for the individual parts.

[0052] In at least one example, the AI control unit 102 employs linear discriminant analysis (LDA), which is a supervised dimensionality reduction technique that seeks to maximize the separation between different classes. In at least one example, LDA aims to find a linear combination of features that best discriminates between fatigued and non-fatigued pilots.

[0053] In at least one example, the AI control unit 102 employs feature scaling to ensure that all features are on the same scale. Feature scaling makes it easier for the machine learning model to learn and improves performance. Feature scaling can include min-max scaling (which is a normalization technique that rescales the features to a specified range), and / or Z-score normalization (which rescales the features with a mean of 0 and a standard deviation of 1).

[0054] In at least one example, the AI control unit 102 uses feature selection techniques to select desired features from the data, thereby reducing noise and improving model performance. For example, Recursive Feature Elimination (RFE) is a backward feature elimination technique that recursively removes the least important features based on their importance in the model until the desired number of features is reached. As another example, SelectKBest is a univariate feature selection technique that selects the top k features (as determined by a scoring function such as chi-squared distribution, ANOVA F-value) based on the relationship between the features and the target variable.

[0055] In at least one example, the AI control unit 102 employs a voting classifier model. For example, the machine learning model 124 can include a voting classifier model. A voting classifier model is an ensemble learning technique that combines the predictions of multiple base classifiers to improve overall performance and reduce the likelihood of overfitting. Voting classifiers can achieve higher accuracy than the individual models they are based on due to the fact that the voting process can balance the individual errors made by the separate models and can handle outliers. Voting classifiers can help mitigate overfitting because they average the results of the individual models, which can reduce the impact of models that overfit the training data. By combining multiple models, voting classifiers can benefit from the strengths of each model and mitigate the weaknesses of each model. Additionally, voting classifiers are flexible because they can combine different machine learning models (including different types of models). For example, a voting classifier can combine a logistic regression model, a decision tree model, and a neural network in the same voting classifier.

[0056] In at least one example, a voting classifier model is trained using a preprocessed and optimized dataset. A combination of base classifiers such as decision trees, support vector machines, and logistic regression is used to build the ensemble model. The model is validated using techniques such as k-fold cross-validation to ensure generality and robustness.

[0057] To evaluate the performance of the voting classifier model, various metrics such as accuracy, precision, recall, F1-score, and the area under the Receiver Operating Characteristic (ROC) curve are used. The main attribute considered in this approach can be the recall value, which is a measure of the proportion of actual positive cases (fatigued pilots) correctly identified by the model. To further optimize the recall value, techniques such as grid search and random search can be used to tune the hyperparameters of the base classifiers and the voting classifier model.

[0058] It has been found that a voting classifier model that trains on an optimized dataset and uses a combination of base classifiers effectively predicts a pilot's readiness state. The voting classifier model achieves high accuracy, precision, recall, and F1-score, indicating its effectiveness in identifying fatigued pilots. The voting classifier model is tailored specifically to the unique needs of pilots, addressing specific flight schedules and rest requirements. This customization results in more accurate readiness state predictions compared to traditional tools that may not fully account for the unique operating conditions. Additionally, the voting classifier model utilizes historical pilot schedules and readiness state self-assessment data to eliminate, minimize, or otherwise reduce the need for invasive physiological data collection. This not only simplifies implementation but also makes the voting classifier model applicable to real-world flight scenarios.

[0059] As described herein, system 100 includes an AI control unit 102 that uses pilot surveys (past and present) and one or more schedules to provide a readiness state assessment tool for evaluating (such as estimating, predicting, etc.) a pilot's readiness state on a given flight. The AI control unit 102 uses one or more machine learning models 124, which can be tailored by using relevant features (e.g., answers to questions posed in the survey) to understand the pilot's readiness state.

[0060] Figure 5 A schematic block diagram of an AI control unit 102 according to an example of the present disclosure is shown. In at least one example, the AI control unit 102 includes at least one processor 400 that communicates with a memory 402. The memory 402 stores instructions 404, received data 406, and generated data 408. Figure 5 The AI control unit 102 shown in is merely exemplary and not restrictive.

[0061] As used herein, the terms "control unit", "central processing unit", "CPU", "computer", etc. can include any processor-based or microprocessor-based system, including systems that use the following: microcontrollers, reduced instruction set computers (RISC), application-specific integrated circuits (ASIC), logic circuits, and any other circuit or processor that includes hardware, software, or a combination thereof capable of performing the functions described herein. This is merely exemplary and is not intended to limit the definition and / or meaning of these terms in any way. For example, the AI control unit 102 can be or include one or more processors configured to control operations as described herein.

[0062] The AI control unit 102 is configured to execute a set of instructions stored in one or more data storage units or elements (such as one or more memories) to process data. For example, the AI control unit 102 may include or be coupled to one or more memories. The data storage unit may also store data or other information as desired or needed. The data storage unit may be in the form of an information source within the processor or a physical storage element.

[0063] The set of instructions may include various commands that direct the AI control unit 102, acting as a processor, to perform specific operations, such as the methods and processes of the various examples of the subject matter described herein. The set of instructions may be in the form of a software program. The software may be in various forms, such as system software or application software. Additionally, the software may be in the form of a collection of separate programs, a subset of programs within a larger program, or a part of a program. The software may also include modular programming in the form of object-oriented programming. The processing of input data by the processor may be in response to a user command, or in response to the result of a previous processing, or in response to a request made by another processor.

[0064] The example diagrams herein may illustrate one or more control or processing units, such as the AI control unit 102. It should be understood that a processing or control unit may represent a circuit, a circuitry, or a part thereof, which may be implemented as hardware having associated instructions (such as software stored on a tangible, non-transitory computer-readable storage medium such as a computer hard drive, ROM, RAM, etc.) for performing the operations described herein. The hardware may include state machine circuitry hard-wired to perform the functions described herein. Optionally, the hardware may include an electronic circuit that includes and / or is connected to one or more logic-based devices such as a microprocessor, a processor, a controller, etc. Optionally, the AI control unit 102 may represent a processing circuitry such as one or more of a field-programmable gate array (FPGA), an application-specific integrated circuit (ASIC), a microprocessor, etc. The circuitry in the various examples may be configured to execute one or more algorithms to perform the functions described herein. One or more algorithms may include aspects of the examples disclosed herein, whether or not explicitly identified in a flowchart or method.

[0065] As used herein, the terms “software” and “firmware” are interchangeable and include any computer program stored in a data storage unit (such as one or more memories) for execution by a computer, the data storage unit including RAM memory, ROM memory, EPROM memory, EEPROM memory, and non-volatile RAM (NVRAM) memory. The above data storage unit types are merely exemplary and are thus not limited to the types of memories that may be used to store computer programs.

[0066] ReferenceFigures 1-5 , examples disclosed in this subject matter provide systems and methods that allow a computing device to quickly and efficiently analyze large amounts of data. For example, the AI control unit 102 can analyze various schedules of a large number of pilots and survey answers from the pilots during a specific time period. Thus, large amounts of data that may not be easily distinguishable by humans are being tracked and analyzed. As described herein, the AI control unit 102 efficiently organizes and / or analyzes large amounts of data. The AI control unit 102 analyzes data in a relatively short period of time to quickly and efficiently evaluate the readiness level of the pilots. Thus, the examples of the present disclosure provide enhanced and efficient functionality, as well as very superior performance, compared to a person analyzing large amounts of data.

[0067] In at least one example, components of the system 100 (such as the AI control unit 102) provide and / or enable a computer system to operate as a special computer system for evaluating the readiness of pilots. The AI control unit 102 improves a standard computing device by determining such information in an efficient and effective manner.

[0068] In at least one example, all or part of the systems and methods described herein can be or otherwise include an artificial intelligence (AI) or machine learning system that can automatically perform operations of the methods also described herein. For example, the AI control unit 102 is an artificial intelligence or machine learning system. These types of systems can be trained and / or self-trained based on external information to repeatedly improve the accuracy of analyzing data to determine a pilot's readiness state. Over time, these systems can be improved by making determinations and communications with increased accuracy and speed, thereby significantly reducing the likelihood of any potential errors. For example, an AI or machine learning system can learn and determine models (such as the machine learning model 124) and associate these models with an assessment of a pilot's readiness state. The AI or machine learning systems described herein can include a number of techniques that are enabled by adaptive prediction capabilities and exhibit at least a degree of autonomous learning to automate and / or enhance pattern detection (e.g., identifying irregularities or regularities in data), customization (e.g., generating or modifying rules to optimize record matching), etc. Feedback from one or more previous analyses of data, integrated data, and / or other such data can be used to train and retrain the system. Based on this feedback, the system can be trained by adjusting one or more parameters, weights, rules, criteria, etc. used in its analysis. This process can be performed using data and integrated data instead of training data and can be repeated multiple times to repeatedly improve the determinations and communications described herein. This training minimizes conflicts and interferences by performing an iterative training algorithm in which an updated data set is used and the system is retrained based on feedback examined prior to the system's most recent training. This provides a robust analysis model that can better evaluate the level of readiness in a cost-effective and efficient manner.

[0069] Figure 6 A perspective front view of an aircraft 106 according to an example of the present disclosure is shown. The aircraft 106 includes a propulsion system 412 that includes, for example, an engine 414. Optionally, the propulsion system 412 can include more engines 414 than shown. The engine 414 is carried by a wing 416 of the aircraft 106. In other examples, the engine 414 can be carried by a fuselage 418 and / or a tail 420. The tail 420 can also support a horizontal stabilizer 422 and a vertical stabilizer 424. The fuselage 418 of the aircraft 106 defines an internal cabin 430 that includes a cockpit or flight deck, one or more work areas (such as a galley, a personal luggage area, etc.), one or more passenger areas (such as first class, business class, and economy class), one or more restrooms, etc. Figure 6 An example of the aircraft 106 is shown. It should be understood that the size, shape, and configuration of the aircraft 106 can be set to be different from Figure 6 that shown.

[0070] In addition, the present disclosure includes examples according to the following terms:

[0071] Clause 1. A system comprising:

[0072] An artificial intelligence (AI) control unit configured to:

[0073] Receive a schedule for a pilot of an aircraft,

[0074] Receive survey answer data from the pilot,

[0075] Analyze the schedule and the survey answer data using one or more machine learning models, and

[0076] Evaluate a readiness level of the pilot based on the schedule and the survey answer data,

[0077] wherein the aircraft is operated according to the readiness level evaluated by the AI control unit.

[0078] Clause 2. The system according to Clause 1, further comprising a user interface in communication with the AI control unit, the user interface including a display in communication with an input device, wherein the AI control unit displays one or more surveys on the display, wherein the pilot inputs answers to questions in the one or more surveys via the input device, and wherein the survey answer data includes the answers.

[0079] Clause 3. The system according to Clause 2, wherein the user interface is on the aircraft.

[0080] Clause 4. The system according to Clause 2 or 3, further comprising a survey database in communication with the AI control unit, wherein the survey database stores survey data including the one or more surveys.

[0081] Clause 5. The system according to any one of Clauses 1-4, further comprising a schedule database in communication with the AI control unit, wherein the schedule database stores schedule data including the schedule.

[0082] Clause 6. The system according to any one of Clauses 1-5, further comprising a model database in communication with the AI control unit, wherein the model database stores the one or more machine learning models.

[0083] Clause 7. The system according to any one of Clauses 1-6, wherein the one or more machine learning models include a plurality of machine learning models.

[0084] Clause 8. The system according to Clause 7, wherein the AI control unit is further configured to select the result of one of the plurality of machine learning models based on the evaluated reliability.

[0085] Clause 9. The system according to any one of Clauses 1-8, wherein the one or more machine learning models include a voting classifier model.

[0086] Clause 10. The system according to any one of Clauses 1-9, wherein the AI control unit is further configured to automatically operate one or more aspects of the aircraft based on the readiness level evaluated by the AI control unit.

[0087] Clause 11. A method, comprising:

[0088] Receiving, by an artificial intelligence (AI) control unit, a schedule for a pilot of an aircraft;

[0089] Receiving, by the AI control unit, survey answer data from the pilot;

[0090] Analyzing, by the AI control unit, the schedule and the survey answer data using one or more machine learning models; and

[0091] Evaluating, by the AI control unit, the readiness level of the pilot based on the use;

[0092] wherein the aircraft is operated according to the readiness level evaluated by the AI control unit.

[0093] Clause 12. The method according to Clause 11, further comprising displaying, by the AI control unit, one or more surveys on a display of a user interface, wherein the pilot inputs answers to questions in the one or more surveys via the input device, and wherein the survey answer data includes the answers.

[0094] Clause 13. The method according to Clause 12, further comprising storing, in a survey database in communication with the AI control unit, survey data including the one or more surveys.

[0095] Clause 14. The method according to any one of Clauses 11-13, further comprising storing, in a schedule database in communication with the AI control unit, schedule data including the schedule.

[0096] Clause 15. The method according to any one of Clauses 11-14, further comprising storing, in a model database in communication with the AI control unit, the one or more machine learning models.

[0097] Clause 16. The method according to any one of Clauses 11-15, wherein the one or more machine learning models include a plurality of machine learning models.

[0098] Clause 17. The method according to Clause 16, further comprising the result of selecting, by the AI control unit, one of the plurality of machine learning models based on the evaluated reliability.

[0099] Clause 18. The method according to any one of Clauses 11-17, wherein the one or more machine learning models include a voting classifier model.

[0100] Clause 19. The method according to any one of Clauses 11-18, further comprising automatically operating one or more aspects of the aircraft based on the readiness level evaluated by the AI control unit.

[0101] Clause 20. A non-transitory computer-readable storage medium comprising executable instructions that, upon execution, cause one or more control units including a processor to perform operations, the operations including:

[0102] Receiving a schedule for a pilot of an aircraft;

[0103] Receiving survey answer data from the pilot;

[0104] Using one or more machine learning models to analyze the schedule and the survey answer data; and

[0105] Evaluating a readiness level of the pilot based on the using,

[0106] wherein the aircraft is operated according to the evaluated readiness level.

[0107] As described herein, examples of the present disclosure provide systems and methods for accurately evaluating a pilot's readiness. In addition, examples of the present disclosure provide customized methods for predicting and mitigating pilot fatigue.

[0108] Although various spatial and orientation terms (such as top, bottom, lower, middle, lateral, horizontal, vertical, front, etc.) may be used to describe examples of the present disclosure, it should be understood that these terms are used only with respect to the orientation shown in the drawings. The orientation may be reversed, rotated, or otherwise changed such that the upper portion becomes the lower portion and vice versa, horizontal becomes vertical, and so on.

[0109] As used herein, a structure, limitation, or element “configured to” perform a task or operation is specifically structured, constructed, or adapted in a manner corresponding to the task or operation. For clarity and to avoid doubt, an object that can only be modified to perform a task or operation is not “configured to” perform the tasks or operations used herein.

[0110] It should be understood that the above description is intended to be illustrative and not restrictive. For example, the above examples (and / or aspects thereof) may be used in combination with each other. Additionally, many modifications may be made to adapt a particular situation or material to the teachings of the various examples of the present disclosure without departing from its scope. While the dimensions and types of materials described herein are intended to define aspects of the various examples of the present disclosure, these examples are in no way restrictive but rather exemplary. After reviewing the above description, many other examples will be apparent to those of ordinary skill in the art. Accordingly, the scope of the various examples of the present disclosure should be determined with reference to the appended claims along with the full scope of equivalents to which these claims are entitled. In the appended claims and the detailed description herein, the terms “including” and “in which” are used as the plain-English equivalents of the respective terms “comprising” and “wherein”. Additionally, the terms “first,” “second,” “third,” etc. are used merely as labels and are not intended to impose numerical requirements on their objects. Moreover, the limitations of the appended claims are not written in means-plus-function format and are not intended to be interpreted under 35 U.S.C. § 112(f) unless and until such claim limitations expressly use the phrase “means for” followed by a statement of function without further structure.

[0111] This written description uses examples to disclose the various examples of the present disclosure (including the best mode), and also enables any person skilled in the art to practice the various examples of the present disclosure (including making and using any device or system and performing any combined method). The patentable scope of the various examples of the present disclosure is defined by the claims and may include other examples that occur to those of ordinary skill in the art. If these other examples have structural elements that are indistinguishable from the literal language of the claims, or if these other examples include equivalent structural elements that are not materially different from the literal language of the claims, then these examples are intended to be within the scope of the claims.

Claims

1. A system (100) comprising: An artificial intelligence control unit, i.e., an AI control unit (102), configured to: Receive a schedule for a pilot of an aircraft (106), Receive survey answer data (128) from the pilot, Analyze the schedule and the survey answer data (128) using one or more machine learning models (124), and Evaluate a readiness level of the pilot based on the schedule and the survey answer data (128), Wherein the aircraft (106) is operated according to the readiness level evaluated by the AI control unit (102).

2. The system (100) according to claim 1, further comprising a user interface (104) in communication with the AI control unit (102), the user interface (104) including a display (110) in communication with an input device (112), wherein the AI control unit (102) displays one or more surveys on the display (110), wherein answers to questions in the one or more surveys are input by the pilot via the input device (122), and wherein the survey answer data (128) includes the answers.

3. The system (100) according to claim 2, wherein the user interface (104) is on the aircraft (106).

4. The system (100) according to claim 2, further comprising a survey database (118) in communication with the AI control unit (102), wherein the survey database (108) stores survey data including the one or more surveys.

5. The system (100) according to claim 1, further comprising a schedule database (116) in communication with the AI control unit (102), wherein the schedule database (115) stores schedule data including the schedule.

6. The system (100) according to claim 1, further comprising a model database (122) in communication with the AI control unit (102), wherein the model database (122) stores the one or more machine learning models (124).

7. The system (100) according to claim 1, wherein, The one or more machine learning models (124) include a plurality of machine learning models (124).

8. The system (100) according to claim 7, wherein, The AI control unit (102) is further configured to select a result of one of the plurality of machine learning models (124) based on an evaluated reliability.

9. The system (100) according to claim 1, wherein, The one or more machine learning models (124) include a voting classifier model.

10. The system (100) according to claim 1, wherein, The AI control unit (102) is further configured to automatically operate one or more aspects of the aircraft (106) based on the readiness level evaluated by the AI control unit (102).

Images