Attraction control system and method using multi-layer perceptron neural networks and machine learning

Abstract

The controllers 30, 48 are used to provide a dynamic ride profile for the ride system 10. The controllers 30, 48 are used to access a dataset indicative of conditions within the environment of the ride system 10 in response to an alert indicative of a fault in the ride system 10 within the environment. The controllers 30, 48 are used to determine a classification from among multiple classifications for the environment based on the dataset and the ride profile of the ride system 10 via a prediction engine 44, 58 trained using supervised machine learning. The controllers 30, 48 are used to determine one or more recommendations via the prediction engine 44, 58 based on the dataset and the classification, and visually or audibly present the classification along with the one or more recommendations.

Description

[Technical Field]

[0001] The present disclosure relates generally to the field of amusement parks. Specifically, embodiments of the present disclosure relate to methods and apparatus used in connection with amusement park games, shows, or rides. [Background technology]

[0002] Amusement parks typically include a variety of attractions that provide guests with unique experiences. For example, amusement parks may include a variety of rides and show performances. As technology continues to advance, the sophistication and complexity of such attractions also increases. For example, some rides may provide visitors with an immersive experience, such as a series of vehicles that may transport passengers through rooms with various features, including audio, video, and special effect features. The increasing sophistication and complexity of modern ride attractions, and the corresponding rising expectations among theme park or amusement park patrons, require amusement parks to provide improved and more creative monitoring and control systems.

[0003] Furthermore, ride systems or show performances may include complex systems such as multi-axis robots, hydraulic and electric motion bases, transportation platforms, and custom assemblies with high-speed operating modes. Conventional techniques employed to monitor ride systems or show performances have very strict operating requirements and very difficult testing methods. It is now recognized that it is desirable to have a dynamic or changing operating environment for ride systems or show performances.

[0004] This section is intended to introduce the reader to various aspects of art that may be related to various aspects of the present technology, which are described and / or claimed below. This discussion is believed to be helpful in providing the reader with background information to facilitate a better understanding of the various aspects of the present disclosure. As such, it should be understood that these statements are to be read in this light, and not as admissions of prior art. Summary of the Invention

[0005] The following provides an overview of some embodiments disclosed herein. It should be understood that these aspects are merely intended to provide the reader with a summary of some embodiments and are not intended to limit the scope of the present disclosure. In fact, the present disclosure may include various aspects that may not be set forth below.

[0006] In one embodiment, the present disclosure provides a controller for a vehicle system that provides a dynamic vehicle profile. The controller includes a processor and a memory. The memory stores instructions that, when executed by the processor, cause the processor to perform operations including: in response to an alert indicating a fault in the vehicle system, accessing a dataset indicative of conditions in an environment of the vehicle system; and determining a classification from a plurality of classifications for the environment based on the dataset and the vehicle profile of the vehicle system via a prediction engine trained using supervised machine learning. The operations further include determining one or more recommendations based on the dataset and the classification via the prediction engine; and visually or audibly presenting the classification along with the one or more recommendations.

[0007] In one embodiment, the present disclosure provides a method for providing a dynamic vehicle profile for a vehicle system. The method includes accessing, via a processor, a dataset indicative of conditions in an environment of the vehicle system in response to an alert indicating a fault in the vehicle system, and determining, via a predictive engine trained using supervised machine learning, a classification from a plurality of classifications for the environment based on the dataset and the vehicle profile of the vehicle system. The method further includes determining, via the predictive engine, one or more recommendations based on the dataset and the classification, and visually or audibly presenting the classification along with the one or more recommendations.

[0008] In one embodiment, the present disclosure provides a ride system that provides a dynamic ride profile. The ride system includes a ride vehicle traveling within an environment, the ride vehicle including an onboard sensor configured to collect vehicle data from the ride vehicle. The ride system includes a dynamic element disposed within the environment and an offboard sensor disposed outside the ride vehicle to collect dynamic element data indicative of at least one aspect of the dynamic element. The ride system includes a controller. The controller includes one or more processors and a memory. The memory stores instructions that, when executed by the one or more processors, cause the one or more processors to perform operations including: accessing a dataset indicative of conditions within the environment of the ride system in response to an alert indicating a fault in the ride system; and determining a classification from a plurality of classifications for the environment via a prediction engine trained using supervised machine learning based on the dataset and the ride profile of the ride system. The operations further include determining one or more recommendations via the prediction engine based on the dataset and the classification, and visually or audibly presenting the classification along with the one or more recommendations.

[0009] These and other features, aspects, and advantages of the present disclosure will be better understood from the following detailed description when read in conjunction with the accompanying drawings, in which like reference characters refer to like elements throughout. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a schematic diagram of an embodiment of a vehicle system according to an embodiment of the present disclosure.

[0011] [Figure 2] 1 is a schematic perspective view of an embodiment of a simulated object within a vehicle profile and associated graphs, in accordance with an embodiment of the present disclosure;

[0012] [Figure 3] 1 is a schematic perspective view of an embodiment of a simulated object outside a vehicle profile and associated logic diagram, according to an embodiment of the present disclosure;

[0013] [Figure 4] FIG. 1 is a block diagram of an embodiment of a vehicle environment in accordance with an embodiment of the present disclosure.

[0014] [Figure 5] FIG. 5 is a block diagram of an embodiment of a ride vehicle within the ride environment of FIG. 4 in accordance with an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0015] One or more specific embodiments are described below. In the interest of brevity in describing these embodiments, not all features of the implementations are described herein. It should be understood that the development of any such implementation, as in any engineering or design project, requires numerous implementation-specific decisions to achieve the developer's particular objectives, including compliance with system- and business-related constraints that may vary from implementation to implementation. Moreover, it should be understood that such a development effort might be complex and time-consuming, but would be a routine undertaking of design, fabrication, and manufacture for those of ordinary skill in the art having the benefit of this disclosure.

[0016] The present embodiment relates to a dynamic system that analyzes and responds to attraction conditions. The dynamic system according to the present embodiment can be layered with a profile evaluation system (e.g., a lookup table for analyzing a ride profile) to facilitate analysis and operation to obtain more desirable results. Specifically, the dynamic system can override an initial evaluation based on the profile evaluation system under certain circumstances. This is particularly useful in attraction systems that allow user input and option selection throughout an attraction sequence, such as rides where passengers can navigate the ride vehicle. For example, a ride profile lookup table (e.g., this lookup table can include multiple tables that change based on passenger size, preferences, and age) can be used to make decisions regarding how certain aspects of the ride system should operate based on ongoing input (e.g., sensor data). As a specific example, if a condition outside the ride profile is detected in a user-guided ride, the profile evaluation system can command a modification based on exceeding the ride profile. However, in some situations, it may be desirable to continue without such a modification. Specifically, if a bird is detected entering the ride path of a vehicle, for example, on a dark ride, the profile assessment system may identify this as an event that deviates from the ride profile, and the initial response may be to stop the ride. However, in accordance with this embodiment, a second, dynamic layer of analysis may be provided that allows the system to override this initial response so that the ride can continue with minor or no modifications so as not to disrupt the ride experience. Further details of operation are provided below.

[0017] When introducing elements of various embodiments of the present disclosure, the articles "a," "an," and "the" are intended to mean the presence of one, two, or more of the element. The terms "comprising," "including," and "having" are intended to be inclusive and mean that there may be additional elements other than the listed elements. Furthermore, references to "one embodiment" or "an embodiment" of the present disclosure are not intended to be interpreted as excluding the existence of additional embodiments that also contain the recited features.

[0018] As used herein, the terms "connected" ("connect," "connection," "connected," "in connection with," and "connecting") can mean "in direct connection with" or "in connection via one or more elements." Additionally, the term "set" is used to mean "one element" or "multiple elements." Additionally, the terms "coupled" ("couple," "coupling," "coupled," "coupled together," and "coupled with") are used to mean "directly coupled together" or "coupled together via one or more elements."

[0019] Also, as used herein, the terms "real time" or "substantially real time" may be used interchangeably and are intended to describe operations (e.g., computational operations) performed without any humanly perceptible interruption therebetween. For example, as used herein, data for the systems described herein may be collected, transmitted, and / or used in control computations in "substantially real time" such that data reading, data transfer, and / or data processing steps occur once per second, once per 0.1 second, once per 0.01 second, or more frequently during operation of the system (e.g., while the system is operating). Also, as used herein, the terms "continuous," "continuously," or "continually" are intended to describe operations performed without any significant interruption. For example, as used herein, control commands may be sent to equipment every 5 minutes, 1 minute, 30 seconds, 15 seconds, 10 seconds, 5 seconds, or more frequently to adjust the operating parameters of the equipment without significantly disrupting the closed-loop control of the particular equipment. Also, as used herein, terms denoting "automatic" or "autonomous," such as "automatic," "automated," and "autonomous," are intended to describe actions that are performed or adapted to be performed by, for example, a computer system (i.e., by the computer system alone and without human intervention).Indeed, although some of the operations described herein may not be explicitly described as being performed continuously and / or automatically in substantially real time during operation of the computer system and / or equipment controlled by the computer system, it will be understood that these operations may in fact be performed continuously and / or automatically in substantially real time during operation of the computer system and / or equipment controlled by the computer system to enhance the functionality of the computer system (e.g., by facilitating faster operational decision-making by not requiring human intervention and by increasing the accuracy of operational decision-making by, for example, eliminating the possibility of human error), as described in further detail herein.

[0020] It is becoming more common for amusement parks to create elaborate environments that include scenery, special effects, audiovisual features, and other media elements that enhance the guest experience. Guests can step into or ride (aboard a ride vehicle) within the environment (e.g., a show performance or attraction) and interact with the environment or have an immersive experience within the environment. Amusement parks may operate under certain standards to maintain a guest experience above a certain level and / or address other issues related to the guest experience at the amusement park (e.g., safety). If the operation of a ride system or the environment does not meet the corresponding standard(s), an alert may be sent to guest and environment control systems to notify guests to exit or reroute from the environment, to shut down one or more elements within the environment, etc. Due to the entertainment nature of amusement parks, these operating criteria may vary for different environments, different ride systems, and / or different guests (e.g., guests of different heights, weights, ages, etc.) to accommodate the needs of individual guests wishing to experience the amusement park.

[0021] According to this embodiment, ride profiles can be used to monitor and / or control the operating criteria of a particular ride. Traditionally, ride profiles for various environments and / or ride systems in an amusement park can be provided by a ride designer or determined based on historical data (e.g., golden profiles (GPs) in look-up tables, CAM profiles) and stored in a database. If the operation of the environment or ride system does not fall within a guest's ride profile, an alert can be sent to the guest or environment for a corresponding response, such as exiting the environment, rerouting the ride, and disabling one or more elements in the environment. For example, in one embodiment, when a guest is present in an environment, sensors (e.g., distributed within the environment, on the ride vehicle, on guest-attached devices) can monitor the guest's state and compare it to values ​​for the corresponding ride profile stored in a database. When a guest's state does not fall within the corresponding ride profile, a trigger event occurs and a corresponding alert can be sent for the guest associated with the ride profile. For example, a trigger event can be a software error that provides an erroneous value to an actuator in the environment, resulting in an inaccurate set point, or an actuator can malfunction and attempt to transition to an unexpected state (e.g., not present in the ride profile provided by the ride designer or not present in the historical data used to determine the ride profile), causing the guest state in the environment to deviate from the corresponding ride profile. Based on the ride profile associated with the guest, a corresponding alert can be sent for the guest associated with the ride profile, requesting the guest to exit the environment or reroute the ride.

[0022] However, because the operational criteria associated with a guest's ride profile are determined based on historical data and / or a ride designer's iterative approach to putting together a desired ride profile, the operational criteria associated with a guest's ride profile under certain circumstances (e.g., associated with a different environment, a different ride vehicle, and / or a desired experience) may not be accurate under some unforeseen circumstances (e.g., not present in the ride profile provided by the ride designer or not present in the historical data used to determine the ride profile). Specifically, due to the entertainment nature of amusement parks, the operational criteria may be different for different environments, different ride systems, and / or different guests (e.g., guests having different heights, weights, ages, etc.). Therefore, to enhance a guest's experience within an amusement park (e.g., within a particular attraction), it may be desirable to reevaluate the guest's condition in real time to determine a dynamic ride profile. For example, in one embodiment, an alert may be sent for the guest to request exit from the environment due to some obstruction within the environment. Alternatively, the prediction engine may determine that the guest does not need to exit the environment and provide a corresponding recommendation to the guest, ensuring that the guest's experience within the environment is not interrupted. Accordingly, the present disclosure relates to determining a guest's dynamic ride profile in an amusement park environment by using a multi-layer perceptron ML (machine learning) model trained with labeled data. The multi-layer perceptron model can be used to solve real-time classification problems with a small number of input features with high accuracy.

[0023] According to this embodiment, a dynamic ride profile may be determined for an amusement park environment or ride system. The dynamic ride profile may be determined based on the characteristics or properties of a particular ride system (including the ride vehicle and the environment). The dynamic ride profile may also be determined based on the state (e.g., physical condition, height, weight, age) and desired experience (e.g., relaxed, thrilling, standard) of an individual guest for the amusement park or a particular attraction (e.g., a particular ride). The dynamic ride profile may be different for different environments or different ride systems within the park for the same guest. The dynamic ride profile may be different for different guests within the same amusement park environment or the same attraction. For example, different guest sizes, ages, or preferences may correspond to different profiles.

[0024] As an example of how this embodiment can operate, consider a scenario in which a guest walks or navigates (e.g., aboard a ride vehicle) through an environment within a system that monitors the guest's ride profile using lookup tables provided by a ride designer or determined based on historical data. One or more sensors (e.g., motion sensors, position sensors, weight sensors, light sensors, sound sensors, image cameras, radio frequency identification (RFID) sensors) can be distributed at one or more locations in the environment to monitor the operation of the environment. There can also be one or more elements within the environment, and these elements can have one or more sensors installed to monitor the operation of the one or more elements. There can also be one or more sensors within the environment that monitor a condition related to the guest. For example, the one or more sensors can be installed on a device (e.g., a handheld device, a wearable device, a tag) attached to the guest or on the ride vehicle in which the guest is seated. A controller can be installed within the device, ride vehicle, or control center to receive data from the sensors monitoring the condition related to the guest, or the operation of the environment, or the operation of one or more elements within the environment. The controller can analyze the data to evaluate a condition associated with a guest, the operation of the environment, or the operation of one or more elements within the environment. The controller can use the data to determine whether a guest's condition meets corresponding criteria associated with the guest's ride profile stored in a lookup table. The controller can send an alert when there is a fault in the environment (e.g., something detected on the ride track, a software error, input from a ride engineer) and when a guest's condition does not meet corresponding criteria associated with the guest's ride profile. The alert can include an action requesting the guest to exit or reroute the environment, or an action instructing the environment's control system to stop or adjust the operation of one or more elements within the environment. However, as discussed above, some of the actions (e.g., exiting the environment) may be unnecessary or undesired by the guest.

[0025] In one embodiment, the controller can receive a trigger signal associated with the alert that can initiate the controller to re-evaluate the data using a prediction engine implementing a multi-layer perceptron model and can be trained to generate a dynamic ride profile using supervised machine learning. Based on the generated dynamic ride profile, the prediction engine can provide a recommendation associated with the alert, which may conflict with the action associated with the alert. For example, the alert includes an action requiring the guest to exit the environment or reroute, whereas the recommendation provided by the prediction engine allows the guest to remain within the environment or route, or allows the environment to continue operating instead of shutting down one or more elements within the environment. For example, the alert can result from some unexpected change in the environment or elements within the environment (e.g., not present in the ride profile provided by the ride designer or not present in the historical data used to determine the ride profile) that can be identified as not meeting the operating criteria in the guest's ride profile. However, these unexpected changes may not affect the guest's overall experience within the environment, and may not require the guest to exit or reroute the environment or to control (e.g., shut down) one or more elements within the environment, and sometimes such requests and controls may result in an undesirable guest experience within the environment. Thus, by employing a predictive engine and using a predictive model to evaluate unexpected changes, guests may have a more lenient or relaxed dynamically changing ride profile than before the predictive model was implemented, and therefore may have a better experience within the amusement park.

[0026] In some embodiments, one or more additional effects may be employed to provide an additional experience related to the recommendation. The controller may operate additional devices or sensors and / or receive additional signals (e.g., by communicating over a network) to interact with the recommendation to enhance the guest's experience and support a particular narrative of the environment. For example, the recommendation may relate to part of a story or scene and may be controlled to operate in a pattern with other attractions in the amusement park. In some embodiments, when an issue is identified that a ride profile does not meet the desired criteria for a particular guest, an alternative experience may be seamlessly provided to provide the guest with the desired experience.

[0027] 1 is a block diagram of a system 10 including a guest 11 traveling within an environment 14 aboard a ride vehicle 12. On the ride vehicle 12, there may be onboard sensors 16 (e.g., motion sensors, position sensors, weight sensors, light sensors, sound sensors, image sensors, cameras, RFID sensors) installed or positioned to collect vehicle data 18 from the ride vehicle 12, which may include information (e.g., image data) related to the guest 11. The onboard sensors 16 may also collect data from the environment 14. Elements 20 may be located within the environment 14, and element sensors 22 (e.g., motion sensors, position sensors, weight sensors, light sensors, sound sensors, image cameras, RFID sensors) may be installed or positioned on the elements 20 to collect element data 24 from the elements 20. Offboard sensors 26 (e.g., motion sensors, position sensors, weight sensors, light sensors, sound sensors, image cameras, RFID sensors) may be installed or positioned within the environment 14 to monitor and collect environmental data 28 from the environment 14. The offboard sensors 26 may collect data related to other operating parameters of the vehicle 12, the elements 20, and / or the environment 14. An onboard controller 30 may be installed on the vehicle 12 to receive and analyze data from the onboard sensors 16, the element sensors 22, and / or the offboard sensors 26.

[0028] The onboard controller 30 may include various types of components that can assist the onboard controller 30 in performing various types of computational tasks and operations. For example, the onboard controller 30 may include a communications component 32, a processor 34, memory 36, storage 38, input / output (I / O) ports 40, a display 42, and a prediction engine 44 that implements a multi-layer perceptron machine learning model. The prediction engine 44 may utilize one or more predictive models to assess the state of the guest 11 within the environment 14 and determine the guest 11's dynamically changing ride profile. Various types of predictive models may be used to analyze data from sensors (e.g., onboard sensors 16, element sensors 22, offboard sensors 26) and generate predicted outputs. The prediction engine 44 is trained using supervised machine learning methods, i.e., the predictive model is trained using training data that includes input data and a desired predicted output (e.g., a labeled dataset). The amount of training data potentially required to train a predictive model may be significant.

[0029] The communications component 32 can be a wireless or wired communications component that can facilitate communications between the on-board controller 30 and various other controllers and devices, such as over a network or the Internet. For example, the communications component 32 can enable the on-board controller 30 to obtain data, such as vehicle data 18, from various data sources. The communications component 32 can use various communications protocols, such as Open Database Connectivity (ODBC), TCP / IP protocol, Distributed Relational Database Architecture (DRDA) protocol, Database Change Protocol (DCP), HTTP protocol, any other suitable current or future protocol, or a combination thereof.

[0030] The processor 34 can process instructions for execution within the onboard controller 30. The processor 34 can include single-threaded processor(s), multi-threaded processor(s), or both. The processor 34 can process instructions stored in memory 36. The processor 34 can also include hardware-based processor(s), each including one or more cores. The processor 34 can include general-purpose processor(s), special-purpose processor(s), or both. The processor 34 can be communicatively coupled to other internal components (such as the communications component 32, storage 38, I / O ports 40, and display 42).

[0031] Memory 36 and storage 38 may be any suitable article of manufacture capable of acting as a medium for storing processor-executable code, data, or the like. These articles of manufacture may represent computer-readable media (e.g., any suitable form of memory or storage) capable of storing processor-executable code used by processor 34 to perform the techniques of the present disclosure. As used herein, an application may include any suitable computer software or program that can be installed on onboard controller 30 and executed by processor 34. Memory 36 and storage 38 may represent non-transitory computer-readable media (e.g., any suitable form of memory or storage) capable of storing processor-executable code used by processor 34 to perform the various techniques described herein. Note that non-transitory merely indicates that the medium is tangible and not a signal.

[0032] The I / O port 40 can be an interface that can be coupled to input devices (e.g., keyboard, mouse), sensors, and other peripheral components such as input / output (I / O) modules. The display 42 can operate as a human-machine interface (HMI) for depicting visualizations related to software or executable code being processed by the processor 34. In one embodiment, the display 42 can be a touch display that can receive input from an operator of the onboard controller 30. The display 42 can be any suitable type of display, such as, for example, a liquid crystal display (LCD), a plasma display, or an organic light-emitting diode (OLED) display. In one embodiment, the display 42 can also be provided with a touch-sensitive mechanism (e.g., a touch screen) that can function as part of the control interface of the onboard controller 30.

[0033] It should be noted that the components described above with respect to on-board controller 30 are exemplary, and on-board controller 30 may include more or fewer components than the illustrated embodiment.

[0034] Additionally, the system 10 may include networking capabilities to facilitate data communication within the system 10 and with external devices. For example, data (e.g., image data, video data, sound data, location data, weight data) collected by sensors (e.g., onboard sensors 16, element sensors 22, offboard sensors 26) may be transmitted to the onboard controller 30 via the network 46. Additionally, external data (e.g., data about a particular user, local weather / news) may be collected from a remote system and transmitted to the onboard controller 30 via the network 46. However, in some embodiments, data collected by the sensors (e.g., onboard sensors 16, element sensors 22, offboard sensors 26) may be transmitted directly to the onboard controller 30. Indeed, according to this embodiment, the onboard controller 30 may communicate with sensors or other devices directly and / or via the network 46.

[0035] The off-board controller 48 may be located at a location other than the vehicle 12 (e.g., an amusement park control center), and the off-board controller 48 may include various types of components that can assist the off-board controller 48 in performing various types of computer tasks and operations. The off-board controller 48 may include components similar to those included in the on-board controller 30. For example, the off-board controller 48 may include a communication component 50, a processor 52, memory 53, storage 54, input / output (I / O) ports 55, a display 56, and a prediction engine 58 similar to the prediction engine 44, etc.

[0036] The off-board controller 48 can receive and analyze data from the on-board sensors 16, the element sensors 22, and / or the off-board sensors 26. Data (e.g., image data, video data, sound data, location data, weight data) collected by the sensors (e.g., the on-board sensors 16, the element sensors 22, the off-board sensors 26) can be transmitted to the off-board controller 48 via the network 46. Additionally, external data collected from remote systems (e.g., data related to a particular guest, local weather / news) can be transmitted to the off-board controller 48 via the network 46. However, in some embodiments, data collected by the sensors (e.g., the on-board sensors 16, the element sensors 22, the off-board sensors 26) can also be transmitted directly to the off-board controller 48. Indeed, according to this embodiment, the off-board controller 48 can communicate with sensors or other devices directly and / or via the network 46. The off-board controller 48 can also communicate with a database 59 that can store information related to attractions and / or particular guests within the amusement park (e.g., the environment 14). The database 59 can be used to store external data.

[0037] In one embodiment, onboard controller 30 and / or off-board controller 48 may analyze data collected by sensors (e.g., onboard sensors 16, element sensors 22, off-board sensors 26) and / or external data (e.g., data about guest 11 on vehicle 12, local weather / news) to assess conditions related to the operation of vehicle 12, environment 14, or elements within environment 14. Onboard controller 30 and / or off-board controller 48 may determine measured conditions of guest 11, including vehicle 12 (or, in some embodiments, not including vehicle 12), and compare the measured conditions to a look-up table 60 stored on onboard controller 30 and / or off-board controller 48 that defines corresponding ride profiles for guest 11 having acceptable operating condition ranges (e.g., taking into account guest physical characteristics and preferences, etc.). The on-board controller 30 and / or off-board controller 48 can issue an alert when there is a fault in the environment 14 (e.g., something detected on the ride track or attraction route, a software error, input from a ride engineer) and the measured conditions of the guest 11 do not meet the corresponding criteria associated with the guest's 11 ride profile. For example, the on-board controller 30 and / or off-board controller 48 can issue an alert for a ride vehicle that is moving too slowly for a particular guest's preferences. The alert can include instructions for the guest 11 (e.g., to change route, fasten restraints, remain seated, or exit the environment 14), control of a particular feature (e.g., stopping one or more elements (e.g., element 20) in the environment 14, increasing the speed of the ride vehicle 12), or guidance for the operator (e.g., adjusting ride parameters, communicating with the guest), etc.

[0038] In FIG. 2 , a schematic diagram 61 shows an object 62 moving through a space having a coordinate axis system XYZ. The object 62 is located within a vehicle profile 63 having allowable operating states / positions between a first boundary surface 64 and a second boundary surface 66. These boundary surfaces are not physical surfaces, but rather represent edges or boundaries of a volume within which the object 62 can or is expected to move. The object 62 is within its vehicle profile 63 (e.g., a golden profile (GP)) when it moves between the first boundary surface 64 and the second boundary surface 66, i.e., within its expected operating range. The object 62 is outside its vehicle profile 63 when it exits the first boundary surface 64 or the second boundary surface 66, as shown in FIG. 3 . As described above, the vehicle profile 63 can change based on conditions. As an example, the physical characteristics of the passengers can be taken into account, so that the first boundary surface 64 and the second boundary surface 66 can be closer to each other for a taller passenger than for a shorter passenger.

[0039] Graph 70 in FIG. 2 illustrates the relative positions of first boundary surface 64, object 62, and second boundary surface 66. Curve 72 of graph 70 illustrates the position of first boundary surface 64 on the XZ plane, curve 74 illustrates the position of object 62 on the XZ plane, and curve 76 illustrates the position of second boundary surface 66 on the XZ plane. In one embodiment, as illustrated by curves 72 and 76 of graph 70, boundaries 64, 66 can vary depending on the positioning of object 62. A vehicle profile 63 of object 62 constrains the position of object 62 to lie between first boundary surface 64 and second boundary surface 66, and thus curve 74 lies between curves 72 and 76, indicating that object 62 is located within an acceptable range. Vehicle profile 63 of object 62 can be determined by implementing a lookup table or utilizing a machine learning algorithm trained by historical data or other identification techniques that use data sets collected by sensors used to detect the state of object 62 and its environment to assess conditions associated with object 62. When the object 62 is traveling outside of its ride profile 63, instructions can be provided by consulting a lookup table or based on the results of a machine learning algorithm. For example, the lookup table can include a mapping between the location of the object 62 and a corresponding notification of its location. The notification of the object 62's location can indicate whether the object 62 is within or outside of its ride profile 63 and can also include corresponding instructions to prevent the object 62 from departing from its ride profile 63. For example, if the object 62 is outside of its ride profile 63, the instructions can include a request or alert for the object 62 to return to its ride profile 63. Although FIG. 2 shows only two boundaries (e.g., boundaries 64, 66), the object 62 can have more than two boundaries that define the limits of its ride profile 63. Additionally, the boundaries can vary for different guests / vehicles due to the individual guest's personal condition (e.g., physical characteristics) and desired experience type, and / or adaptation to the vehicle features and characteristics of the vehicle 12, etc.

[0040] FIG. 3 illustrates an embodiment in which object 62 has moved outside its vehicle profile 63 (or vehicle envelope). Specifically, diagram 80 shows that object 62 has moved beyond first boundary surface 64. According to this embodiment, such movement beyond vehicle profile 63 results in the generation of a corresponding command (e.g., via onboard controller 30 or offboard controller 48). The command may reposition object 62 within first boundary surface 64, present a request to a user or motion operator to reposition object 62, or provide an alert indicating that object 62 is outside vehicle profile 63. The vehicle profile 63 of object 62 is an indicator and guide for keeping object 62 within a desired range of movement when moving or repositioning object 62 to avoid issues (e.g., potential collisions) that could result in downtime or other undesirable events. In other words, the ride profile 63 of the object 62 can be used to keep the object 62 within a particular space for a given environment within the amusement park, or to avoid undesirable events due to walls and / or background props encroaching on the ride envelope of the object 62. Again, the boundary surface can move with the object 62, as shown in graph 70 of FIG.

[0041] However, unexpected changes may occur in the object 62 and / or the surrounding environment (e.g., environment 14), such as an obstacle (e.g., something detected near or in the path of the object 62, a software error, an input related to the object 62), etc. The lookup table may not have a corresponding entry for the unexpected change (e.g., because the unexpected change was not accounted for in the lookup table). Therefore, it is desirable to have a dynamically changing vehicle profile in the time domain of the object 62.

[0042] According to this embodiment, a prediction engine may be employed that implements a multi-layer perceptron machine learning model that uses supervised machine learning to rapidly make one or more decisions based on unexpected changes, either separately from or as a supplement to the lookup table. The prediction engine may utilize one or more predictive models to determine a dynamically changing vehicle profile of the object 62 and determine a classification from multiple classifications of the environment based on the dynamically changing vehicle profile of the object 62. For example, the multiple classifications may include some classifications that indicate that the unexpected changes may cause the object 62 to fall outside a threshold location range and some classifications that indicate that the unexpected changes may not cause the object 62 to fall outside the threshold location range. The prediction engine may provide recommendations based on the classifications.

[0043] Diagram 82 in FIG. 3 illustrates predictive modeling performed by one or more controllers (e.g., onboard controller 30) of system 10 according to an embodiment. Predictive modeling generally refers to computer-based techniques, algorithms, systems, and the like, used to extract information from data and build models capable of predicting an output from a given input. For example, predictive modeling can predict future outcomes that can be expected to occur within a short period of time from the time the measurements were taken, given several measurements (e.g., from sensors). In diagram 82, block diagram 84 illustrates using a single input 85 to obtain a single output 87 via logic 86. Also in diagram 82, block diagram 88 illustrates using two inputs 85, which may be related via logic 86, to obtain two outputs 87. Various types of predictive models can be used to analyze data and generate predicted outputs. In supervised machine learning methods, a predictive model is trained using training data that includes input data and desired predicted outputs (e.g., a labeled dataset). The amount of training data potentially required to train a predictive model can be large.

[0044] Referring again to FIG. 1 , vehicle 12 moves through environment 14 similarly to object 62 moving along X, Y, and Z coordinates in FIG. 2 . Unexpected changes (i.e., changes not included in lookup table 60) may occur in guest 11, vehicle 12, and environment 14 (e.g., elements 20). Examples of unexpected changes may include mechanical wear (e.g., mechanical wear of vehicle 12, elements 20, or other elements within environment 14), uncontrollable factors (e.g., weather, political or economic environment, earthquakes, hurricanes), unexpected behavior of vehicle 12, and / or any combination thereof. When such unexpected changes are detected, system 10 may generate commands, alerts, and / or initiate actions. For example, onboard controller 30 or offboard controller 48, or other controllers associated with system 10, may operate to shut down one or more elements (e.g., elements 20) within environment 14 in response to detecting an unexpected event. As another example, the vehicle 12 may receive an alert indicating an unexpected change in the system 10 (e.g., an obstruction on the ride route). Accordingly, such an alert may include a request for the guest 11 on the vehicle 12 to change to a different route or to exit the environment 14.

[0045] Data and / or instructions related to the unexpected event may be generated by any of various aspects of the system 10. For example, the vehicle 12 may receive an alert or instruction directly or via the network 46 from the onboard controller 30, a feature of the environment 14, or an offboard controller 48. The alert may be provided using various methods, such as via light, sound, vibration, and images. For example, the system 10 may provide the alert, instruction, or other communication using speakers, lights, haptics, and the like, on the vehicle 12, in the environment 14, or on a device (e.g., a wearable device, tag) carried by the guest 11.

[0046] When an unexpected event is detected, the onboard controller 30 and / or the offboard controller 48 may employ one or more prediction engines to analyze the unexpected change using data collected by sensors (e.g., onboard sensors 16, element sensors 22, offboard sensors 26) and / or external data (e.g., data about the guest 11 on the vehicle 12, local weather / news) to evaluate conditions related to the operation of the vehicle 12, the environment 14, or the operation of elements within the environment 14. The prediction engine may evaluate the unexpected change using one or more predictive models, such as those described above, and determine a more lenient or relaxed dynamically changing ride profile for the guest 11 compared to before the predictive model was implemented. The prediction engine may determine a classification of the environment 14 for the unexpected change from multiple classifications based on the dynamically changing ride profile of the guest 11. For example, the plurality of classifications may include some classifications with a conclusive status, indicating that the unexpected change may affect the guest 11's experience within the environment 14 and that certain action should be taken to avoid an undesirable experience for the guest 11 due to the unexpected environmental change. The plurality of classifications may also include some classifications with an optional status, indicating that the unexpected change may not affect the guest 11's experience within the environment 14 and that taking any action related to the guest 11 regarding the unexpected change is considered unnecessary or optional. The prediction engine may provide recommendations to the system 10 based on the classifications. For example, some unexpected changes may be predicted to not affect the guest's overall experience within the environment 14 (e.g., classifications with optional status). If such a prediction prevails (e.g., selected or determined to be likely), issuing a command or request that may detract from the guest's experience (e.g., asking the guest 11 to exit or reroute the environment 14 or shutting down one or more elements within the environment 14) may be avoided.For example, in one embodiment, an alert may be sent for guest 11 that indicates a fault within environment 14 may require guest 11 to exit environment 14, but the prediction engine may determine that exiting environment 14 is not necessary and provide a corresponding recommendation to guest 11 (e.g., giving guest 11 the option to remain within environment 14). On the other hand, some unexpected changes may be predicted to affect the guest's overall experience within environment 14 (e.g., classifications with deterministic status). If such a prediction prevails (e.g., is selected or determined to be likely), a command or request may be necessary that could detract from the guest's experience (e.g., requesting guest 11 to exit or reroute from environment 14, or shutting down one or more elements within environment 14), and the prediction engine may provide a corresponding recommendation to guest 11. For example, in the example described above, the prediction engine may provide a recommendation that confirms the request in the alert to require guest 11 to exit environment 14.

[0047] As a result, the dynamically changing ride profile of the guest 11 may be more lenient or relaxed compared to before the predictive model was implemented, and the recommendation may include instructions to allow an action that was not permitted or prohibited in the alert. In some embodiments, a recommendation may be sent to the guest 11 in the vehicle 12, and the guest 11 may select an action recommended in the recommendation (e.g., continuing the route rather than changing the route, or remaining in the environment 14 rather than exiting), which may not have been permitted in the alert. In some embodiments, one or more actions in the recommendation may be automatically executed by the system 10 (e.g., the alert may request the system 10 to disable one or more elements in the environment 14, while the recommendation may include an action requesting the system 10 not to disable one or more elements in the environment 14). By employing a predictive engine and using a predictive model to evaluate unexpected changes, the guest 11 may have a dynamically changing ride profile that is more lenient or relaxed compared to before the predictive model was implemented. The guest 11 may then have a better experience within the amusement park.

[0048] 4 and 5 illustrate examples of the use of a prediction engine within an environment 14. FIG. 4 is a block diagram of one embodiment of an environment 14 having one or more attractions separated into different rooms. For example, FIG. 4 illustrates six rooms within the environment 14. Specifically, the environment 14 of FIG. 4 includes a first room 14a, a second room 14b, a third room 14c, a fourth room 14d, a fifth room 14e, and a sixth room 14f. These six rooms may be collectively referred to as rooms 14a-f. Additionally, in the illustrated embodiment, the environment 14 includes an exit 15, which may be a parking space for the vehicle 12. The rooms 14a-f within the environment 14 may be connected, allowing the vehicle 12 to move from one room to another. An off-board controller 48 may collect data from sensors distributed throughout the rooms 14a-f. Each room may be connected to one or more other rooms within the environment 14. In one embodiment, at least one room in the environment 14 may be connected to other locations outside the environment 14 .

[0049] 5 is a perspective view of a guest 11 in a vehicle 12 passing through a room 14a being monitored and analyzed using a predictive model, according to an embodiment. In the illustrated embodiment, element 20 comprises a robotic arm operable to move around within environment 14 for entertainment purposes. An item 92 (e.g., an unexpected bird) can fly into environment 14 and land on element 20, causing an unexpected change in element 20. Onboard sensor 16, element sensor 22, and / or off-board sensor 26 can detect item 92 and send relevant information to onboard controller 30 and / or off-board controller 48.

[0050] Guest 11 may have a ride profile associated with room 14a, and the unexpected change may be identified as interfering with the guest's ride profile based on a lookup table associated with room 14a. In response to identifying this interference, on-board controller 30 and / or off-board controller 48 may issue an alert or command that may be sent to room 14a (e.g., room 14a may have speaker 94), vehicle 12, and / or guest 11 requesting guest 11 to exit room 14a. In conjunction with this alert, the onboard controller 30 and / or the offboard controller 48 may employ one or more prediction engines (e.g., prediction engines 44 and / or 58) to analyze unexpected changes using data collected by sensors (e.g., onboard sensors 16, element sensors 22, offboard sensors 26) and / or external data (e.g., data about the guest 11 in the vehicle 12, local weather / news) to assess conditions related to the vehicle 12, the operation of the room 14a, or the operation of the elements 20 within the room 14a. The prediction engines 44 and / or 58 may use one or more predictive models, such as those described above, to determine a dynamically changing ride profile of the guest 11 in the room 14a. The prediction engines 44 and / or 58 may determine one of multiple classifications for the guest 11's condition based on the dynamically changing ride profile of the guest 11 in the room 14a. For example, the plurality of classifications may include some classifications indicating that the unexpected change may affect the experience of guest 11 in room 14a and some classifications indicating that the unexpected change is not likely or substantially not likely to affect the experience of guest 11. The plurality of classifications may also include some classifications indicating that the unexpected change may affect the experience of guest 11 in room 14a but is not likely to affect the experience of guest 11 in other rooms (e.g., room 14b, room 14c, room 14d, room 14f) of environment 14. The prediction engine may provide recommendations to system 10 based on the classifications.For example, some unexpected changes may be predicted to have no or substantially no likelihood of affecting the guest's overall experience in room 14 a, and thus, if such a prediction prevails (e.g., is selected or determined to be likely), commands or requests that may detract from the guest's experience (e.g., asking guest 11 to leave room 14 a) may be avoided.

[0051] In some embodiments, the on-board controller 30 may provide recommendations more quickly than the off-board controller 48 because the on-board controller 30 requires less time to transmit data. As a result, the on-board controller 30 may be used to provide recommendations when timing is important and real-time data analysis would be beneficial. For example, when an item 92 has not been identified as creating an undesirable experience for guest 11, it may take less time for the on-board controller 30 than the off-board controller 48 to provide a recommendation suggesting that guest 11 not exit room 14a even though an alert requests guest 11 to do so. Thus, in the above example, guest 11's experience in room 14a may not be interrupted, or may not be interrupted for as long. Meanwhile, the off-board controller 48 may have access to more data and therefore may provide more recommendation options or more accurate recommendations. For example, the off-board controller 48 can receive data from all rooms in the environment 14 (i.e., rooms 14a, 14b, 14c, 14d, 14e, and 14f) and analyze the overall situation of the environment 14 to obtain more and more accurate prediction results. Thus, the off-board controller 48 can be used to provide more recommendation options or more accurate recommendations. For example, because the off-board controller 48 can collect data from all rooms in the environment 14, it can provide a recommendation that the guest 11 move to another room (e.g., room 14b, 14c, 14d, 14e, or 14f) rather than exiting the environment 14 (e.g., returning to the exit 15).

[0052] While only certain features of the invention have been illustrated and described herein, many modifications and changes will occur to those skilled in the art and it is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the invention.

[0053] The technology shown and claimed herein refers to and applies to tangible objects and specific examples of a practical nature that will materially improve the art, and thus are not abstract, intangible, or purely theoretical. Furthermore, if any claim appended at the end of this specification contains one or more elements designated as "means for [performing] ... [function]" or "step for [performing] ... [function]," such elements are to be construed pursuant to 35 U.S.C. 112(f). Conversely, for any claim containing elements designated in any other manner, such elements are not to be construed pursuant to 35 U.S.C. 112(f).

Claims

1. 1. A controller for a vehicle system, comprising: a processor; a memory for storing instructions; wherein the instructions, when executed by the processor, accessing a dataset indicative of conditions within an environment of the vehicle system in response to an alert indicative of a fault in the vehicle system; determining a classification from a plurality of classifications for the environment based on the dataset and a vehicle profile or envelope via a prediction engine trained using supervised machine learning; determining one or more recommendations via the prediction engine based on the dataset and the classification; visually or audibly presenting the classification along with the one or more recommendations; causing the controller to perform operations including: controller.

2. the prediction engine is trained using the supervised machine learning based on a plurality of labeled data, the plurality of labeled data being related to the environment; The controller of claim 1 .

3. the plurality of classifications includes discretionary status and definitive status; The controller of claim 1 .

4. at least one of the one or more recommendations is prohibited based on the alert indicating the fault, the fault being an environmental fault of the vehicle system; The controller of claim 1 .

5. the processor is located on a vehicle of the vehicle system; The controller of claim 1 .

6. the processor is configured to access the data set via a sensor set; The controller of claim 1 .

7. the sensor set is disposed within the environment; The controller of claim 6.

8. 1. A method of controlling a vehicle system, comprising: accessing, via a processor, a dataset indicative of conditions within the environment in response to an indication of a fault in the vehicle system; determining a classification from a plurality of classifications for the environment based on the dataset and a vehicle envelope of the vehicle system via a prediction engine trained using supervised machine learning; determining one or more recommendations via the prediction engine based on the dataset and the classification; audibly or visually presenting the classification along with the one or more recommendations; A method comprising:

9. the prediction engine is trained using the supervised machine learning based on a plurality of labeled data, the plurality of labeled data being related to the environment; The method of claim 8.

10. the plurality of classifications includes discretionary status and definitive status; The method of claim 8.

11. at least one of the one or more recommendations is prohibited based on an indication that the fault in the vehicle system is an environmental fault; The method of claim 8.

12. the processor is located on a vehicle of the vehicle system; The method of claim 8.

13. the processor is configured to access the data set via a sensor set; The method of claim 8.

14. the sensor set is disposed within the environment; The method of claim 13.

15. 1. A vehicle system comprising: a ride vehicle configured to move within an environment, the ride vehicle including an on-board sensor configured to collect vehicle data from the ride vehicle; dynamic elements positioned within the environment; an off-board sensor disposed external to the ride vehicle and configured to collect dynamic element data indicative of at least one aspect of the dynamic element; a controller including one or more processors and a memory for storing instructions; wherein the instructions, when executed by the processor, accessing a data set including the vehicle data and the dynamic element data from the on-board and off-board sensors in response to an indication of a fault in the vehicle system; determining a classification from a plurality of classifications for the environment based on the dataset via a prediction engine trained using supervised machine learning and a vehicle envelope of the vehicle system; determining one or more recommendations via the prediction engine based on the dataset and the classification; presenting the classification along with one or more recommendations; causing the controller to perform operations including: Vehicle system.

16. the on-board sensors are configured to collect environmental data from the environment of the vehicle system, and the data set includes the environmental data; 16. The vehicle system of claim 15.

17. The dynamic element includes a path controller, an automatic figure, a maintenance device, or a combination thereof.

16. The vehicle system of claim 15.

18. at least one of the one or more recommendations is prohibited based on the indication of the fault in the vehicle system; 16. The vehicle system of claim 15.

19. the off-board sensor is integrated with the dynamic element; 16. The vehicle system of claim 15.

20. the dynamic elements include an automated figure configured to interact with the ride vehicle during operation of the ride system; 20. The vehicle system of claim 19.

Images