Straddle-type vehicle-related CHL information output system

Abstract

The purpose of the present invention is to construct a system capable of improving the well-being of a rider and to reduce loads on hardware resources constituting the system by effectively and efficiently enhancing cooperative interaction between a straddle-type vehicle and the rider. In the system, by using at least one type of straddle-type vehicle-related data acquired over time from at least one input device in a data acquisition process, straddle-type vehicle-related prompt data which can be interpreted linguistically by a large language model (LLM) and has a reduced amount of data is generated. The straddle-type vehicle-related prompt data is supplied to the LLM, and running straddle-type vehicle linguistic interpretation data is acquired from the LLM. By using the acquired running straddle-type vehicle linguistic interpretation data, CHL information indicating a linguistic interpretation result regarding a running state of the straddle-type vehicle is outputted from at least one output device.

Description

Straddle Vehicle-related CHL Information Output System

[0001] The present invention relates to a straddle vehicle-related CHL (Communicative Human Language) information output system.

[0002] Patent Document 1 discloses an information output device for a straddle vehicle. The information output device includes a receiving unit that receives input information input by a rider, a learning unit that learns an input tendency, which is a tendency of the driver's input information, based on the input information received by the receiving unit, a creating unit that creates output information based on the learning content learned by the learning unit, and an output unit that causes the output information created by the creating unit to be output as natural language voice from a voice device. The information output device can output, for example, an evaluation or a proposal for vehicle body settings or driving operations in natural language according to the input tendency. In addition, the information output device learns the generation rule of voice information based on the rider's reaction to the voice information. Furthermore, the information output device can learn the rule for setting the output timing based on the driver's reaction to the output timing of the voice information.

[0003] Japanese Patent No. 6860553

[0004] An object of the present invention is to construct a system that can effectively and efficiently enhance the sense of cooperation between a straddle vehicle and a rider, improve the well-being of the rider, and reduce the load on the hardware resources constituting the system.

[0005] In view of the above problems, the present inventor has conducted intensive studies and obtained the following findings.

[0006] The saddle-type vehicle referred to in the present invention is configured to be able to turn in a lean position. That is, the saddle-type vehicle is equipped with handlebars, and is configured so that the rider leans the vehicle body and turns by shifting his or her weight while gripping the handlebars with both hands. As a result, when the saddle-type vehicle turns, the rider leans together with the vehicle body toward the inside of the curve. In this way, when turning, the rider uses his or her entire body to move the saddle-type vehicle, and the saddle-type vehicle is configured to lean together with the rider in response to the rider's movement. Turning such a saddle-type vehicle can feel to the rider like a collaborative effort between the rider and the saddle-type vehicle. Therefore, compared to vehicles other than saddle-type vehicles that can turn in a lean position, such as ordinary four-wheeled automobiles (hereinafter also referred to as non-lean vehicles), the saddle-type vehicle is a vehicle that is more likely to create a sense of unity with the rider. The stronger the sense of unity the rider feels with the saddle-type vehicle, the more enjoyable the rider's experience of driving the saddle-type vehicle.

[0007] As such, compared to non-lean vehicles, saddle-ride vehicles are vehicles that tend to make the act of driving itself more enjoyable, and by enhancing the sense of collaboration between the saddle-ride vehicle and the rider, the rider's well-being can be improved. The sense of collaboration between the saddle-ride vehicle and the rider can be enhanced not only by the response of the saddle-ride vehicle (vehicle body) to the rider's operations, but also by the provision of information from the saddle-ride vehicle to the rider. However, while providing information from the saddle-ride vehicle to the rider can dramatically enhance the sense of collaboration if done well, there is a risk that the sense of collaboration may be significantly impaired if done poorly. Even if the saddle-ride vehicle responds well to the rider's operations, if the saddle-ride vehicle provides information to the rider poorly, the rider's impression of the saddle-ride vehicle will be negatively affected. Therefore, it is important to consider how information is provided from the saddle-ride vehicle to the rider.

[0008] In order to successfully provide information from a saddle-type vehicle to a rider, the rider's preferences, emotions, personality, and situation must be acquired in greater detail, increasing the number of factors that must be considered as input information. In addition, the output information, such as the content, transmission method, and transmission timing of the information, must also be set in greater detail, increasing the number of factors that must be considered. In other words, in order to achieve a higher level of interaction between the saddle-type vehicle and the rider, the number of factors that must be considered increases.

[0009] Furthermore, as described above, the saddle-type vehicle referred to in the present invention is configured to be able to turn in a lean position. Therefore, in addition to operating the accelerator, brakes, and steering, the rider must also shift his or her own weight to perform a turning operation in conjunction with steering. This weight shift and the resulting vehicle lean are not present in a non-lean vehicle, so compared to a non-lean vehicle, more information can be obtained from a saddle-type vehicle that can turn in a lean position. This can also be said to mean that there is more information available to achieve a higher level of interaction between the saddle-type vehicle and the rider. A notable example is a manual saddle-type vehicle in which different operations are performed with both hands and feet, as follows: Left hand: operates the clutch lever Right hand: operates the throttle (accelerator) and front brake lever Left foot: operates the gear shift pedal (however, gear shifting may also be performed using a manual lever) Right foot: operates the rear brake As the rider shifts his or her own weight, the operating parts corresponding to each of the hands and feet can be operated at different times for different vehicle operations, further increasing the amount of information that can be obtained. Saddle-type vehicles that can turn in a lean position are not limited to this example, but regardless of the type of vehicle, saddle-type vehicles that can turn in a lean position tend to have a greater variety of information that can be obtained compared to non-lean vehicles.

[0010] Furthermore, in a saddle-type vehicle that can turn in a lean position, it is difficult to obtain much information actively input from the rider other than driving operations while driving. Information actively input other than driving operations is likely to reflect the rider's emotions, which is important from the perspective of interaction with the saddle-type vehicle. However, in a saddle-type vehicle that can turn in a lean position, many parts of the body (both hands, both legs, and weight shift) are required for driving operations, and many working areas of the brain are used for driving, leaving little working area available for operations other than driving, making it difficult for sudden changes in emotions to occur. In addition, the installation space for operating parts other than driving operations (such as plywood panels) is smaller than in a non-lean vehicle. Therefore, it is difficult for the rider to actively input a large amount and variety of information other than driving operations while driving. As a result, while the variety and amount of information that can be obtained is large, the variety and amount of information actively input from the rider other than driving operations is limited.

[0011] As described above, in a saddle-type vehicle that can turn in a lean position, there are many factors that must be considered and a large amount of information that can be used to achieve a high level of interaction between the saddle-type vehicle and the rider. To achieve this high level of interaction between the saddle-type vehicle and the rider, it is necessary to take these factors and information into account and provide outputs that are more easily accepted by the rider in a more timely manner. This presents a unique challenge for high-level interaction between a saddle-type vehicle that can turn in a lean position and the rider.

[0012] As the above-mentioned trends of "increasing factors to be considered," "increasing amount of available information," and "higher demands for timeliness and appropriateness of output" become stronger, the load on the hardware resources required for the information output system increases, and more hardware resources are required. In addition, since a large amount of learning data needs to be collected, preparation takes a long time. The device disclosed in Patent Document 1 also has a similar problem. That is, the device disclosed in Patent Document 1 has the problem of requiring enormous hardware resources and learning data in order to achieve a high level of interaction between the saddle-riding vehicle and the rider that can enhance the sense of cooperation between the saddle-riding vehicle and the rider.

[0013] The present invention has been completed based on the above findings. The present invention can provide the following saddle-ride type vehicle related CHL information output system.

[0014] (1) A straddle-type vehicle related CHL information output system, the straddle-type vehicle related CHL information output system including: at least one memory; at least one processor capable of communicating with an LLM (Large Scale Language Model), connected to the memory, and configured to execute at least one program stored in the memory; at least one input device provided on the straddle-type vehicle or on equipment or carried by a rider of the straddle-type vehicle, configured to acquire or generate at least one type of straddle-type vehicle related data that is related to the traveling state of the straddle-type vehicle and changes as the straddle-type vehicle travels, and to supply the at least one type of straddle-type vehicle related data to the at least one processor; and at least one output device configured to output CHL information, the at least one program comprising: a data acquisition process for acquiring the at least one type of straddle-type vehicle related data from the at least one input device over time while the straddle-type vehicle is traveling; a prompt generation process that generates saddle-riding-type vehicle-related prompt data that can be linguistically interpreted by the LLM and has a reduced amount of data, using the at least one type of saddle-riding-type vehicle-related data acquired over time by the data acquisition process; a linguistic interpretation process that supplies the saddle-riding-type vehicle-related prompt data to the LLM and acquires from the LLM linguistically interpreted data for the saddle-riding-type vehicle that has been generated by the linguistic interpretation by the LLM; and an output process that outputs CHL information indicating the linguistic interpretation result for the driving state of the saddle-riding type vehicle from the at least one output device, using the linguistically interpreted data for the saddle-riding type vehicle acquired from the LLM.

[0015] In the system (1), an LLM is used to provide information from the saddle-type vehicle to the rider in order to effectively and efficiently enhance the sense of cooperation between the saddle-type vehicle and the rider. Additionally, the system (1) has a configuration suitable for using an LLM to provide information from the saddle-type vehicle to the rider. That is, in the system (1), saddle-type vehicle-related prompt data that can be linguistically interpreted by the LLM and has a reduced data volume is generated using at least one type of saddle-type vehicle-related data acquired over time from at least one input device through a data acquisition process. The saddle-type vehicle-related prompt data is supplied to the LLM, and traveling saddle-type vehicle linguistic interpretation data is acquired from the LLM. Furthermore, in the system (1), CHL information indicating the linguistic interpretation result of the traveling state of the saddle-type vehicle is output from at least one output device using the acquired traveling saddle-type vehicle linguistic interpretation data. This makes it possible to effectively and efficiently realize high-level interaction between the saddle-type vehicle and the rider. As a result, the system (1) effectively and efficiently enhances the sense of collaboration between the saddle-type vehicle and the rider, thereby improving the rider's well-being and reducing the load on hardware resources. In other words, assuming the same hardware resources are used, more advanced processing is possible, thereby realizing a higher level of interaction. The system (1) is able to address at a high level the "increase in factors to be considered," "increase in available information," and "increased demands for timeliness and appropriateness of output" while reducing the load on hardware resources, thereby resolving the aforementioned challenges specific to high-level interaction between a saddle-type vehicle capable of turning in a lean position and the rider. From the perspective of achieving the above-mentioned effect (achieving both "reduced load on hardware resources" and "high-level interaction"), generating linguistically interpretable prompt data and supplying it to the LLM while reducing the amount of saddle-type vehicle-related data acquired over time has significant technical significance.

[0016] The straddle-type vehicle-related CHL information output system includes at least one input device, at least one memory, at least one processor, and at least one output device. These devices are communicatively connected to each other. These devices may be provided in a single device to form a centralized processing system, or may be provided in multiple devices communicatively connected to each other to form a distributed processing system. The at least one input device is provided in the straddle-type vehicle or in equipment or equipment carried by the rider. The at least one memory may be provided in the straddle-type vehicle, in equipment or equipment carried by the rider, a user terminal, or a server. The at least one processor may be provided in the straddle-type vehicle, in equipment or equipment carried by the rider, a user terminal, or a server. The at least one output device may be provided in the straddle-type vehicle, in equipment or equipment carried by the rider, or a user terminal. The straddle-type vehicle-related CHL information output system according to the present invention can employ, for example, any of the system configurations (a) to (x) shown in FIGS. 2 and 3 . FIG. 2 shows a device in which an input device (I), a processor (P), a memory (M), and an output device (O) are provided in the system configuration. FIG. 3 shows a device in which an input device, a processor, and an output device are provided. For ease of understanding, FIGS. 2 and 3 use the same reference numerals as those used in the embodiments described below, but these reference numerals are not intended to limit the scope of the present invention. In FIG. 2, the memory is provided in the device in which the processor is provided. In FIG. 3, the memory is also provided in the device in which the processor is provided, although not shown. However, the present invention is not limited to this example. The memory may be provided in a device different from the device in which the processor is provided (e.g., a device in which the input device and / or output device is provided). Furthermore, there may be multiple input devices, output devices, memories, and processors. Therefore, if a single system has multiple devices, the single system may satisfy the requirements of multiple system configurations (a) to (x) shown in FIGS. 2 and 3.

[0017] Examples of system configurations will be described. As a specific example, a smartphone, which corresponds to a rider's personal belongings, may constitute the system because it includes at least one input device, at least one output device, at least one memory, and at least one processor (see FIGS. 2( n) and 3( n)). As another specific example, a sensor provided on the saddle-riding vehicle may correspond to at least one input device, a meter on the saddle-riding vehicle may correspond to at least one output device, and a server capable of communicating with the saddle-riding vehicle may include at least one memory and at least one processor (see FIGS. 2( d) and 3( d)). In one example, the smartphone corresponds to a saddle-riding vehicle-related IPG (Interactive Prompt Generator) that generates saddle-riding vehicle-related prompt data to be supplied to the LLM and acquires CHL information from the LLM. In a second example, the server corresponds to the saddle-riding vehicle-related IPG. In the saddle-riding vehicle-related CHL information output system, all processes may be executed by a single processor, or each process may be executed by a different processor. The system may include: at least one input device and at least one output device, each provided in the saddle-riding type vehicle or in equipment or carried by the rider; and at least one processor and at least one memory, each provided in the saddle-riding type vehicle, the equipment or carried by the rider, a server, or a user terminal ((a) to (p) of FIGS. 2 and 3). The system may include: at least one input device, each provided in the saddle-riding type vehicle or in equipment or carried by the rider; and at least one processor and at least one memory, each provided in the saddle-riding type vehicle, the equipment or carried by the rider, a server, or a user terminal ((q) to (x) of FIGS. 2 and 3). The hardware configuration of the saddle-riding type vehicle-related CHL information output system is not particularly limited.

[0018] As shown in FIGS. 2 and 3 , the input device may be provided in the saddle-riding vehicle, or in equipment or portable equipment of the rider. The input device acquires or generates at least one type of saddle-riding vehicle-related data. The saddle-riding vehicle-related data indicates saddle-riding vehicle-related parameters that are related to the traveling state of the saddle-riding vehicle and change as the saddle-riding vehicle travels. An example of an input device is a sensor that detects the saddle-riding vehicle-related parameters. The input device may function independently and have a communication function so that it can independently output the saddle-riding vehicle-related data. The input device that functions independently is not particularly limited, and examples include a GPS (Global Positioning System) receiver and a GNSS (Global Navigation Satellite System) receiver. The input device may be configured by combining a detection device and a control device. In this case, examples of the detection device include a GPS module, a GNSS module, and an IMU (Inertial Measurement Unit). Examples of the control device include an ECU (Electronic Control Unit). In this way, the input device is composed of a single device or multiple devices. Examples of the input device and the saddle-type vehicle-related data acquired or generated by the input device are not particularly limited, but include the following examples: GPS receiver: driving position information GNSS receiver: driving position information IMU (6-axis sensor): acceleration in three axes (forward / backward, left / right, and up / down) of the saddle-type vehicle while in motion, as well as angular velocity in three axes (pitch, roll, and yaw) Throttle opening sensor: throttle opening Brake position sensor: brake operation ON / OFF or operation amount (operation pressure) Clutch sensor: clutch operation ON / OFF or operation amount Gear position sensor: gear position Steering angle sensor: steering angle Steering torque sensor: steering torque Vehicle speed sensor: vehicle speed Wheel speed sensor: rotational speed of front or rear wheels Engine rotation speed sensor: engine rotation speed

[0019] Processors include central processing units (CPUs), microprocessors, general-purpose processors, digital signal processors (DSPs), graphics processing units (GPUs), controllers, microcontrollers, programmable logic devices (PLDs), field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), etc. Multiple processors can be configured using any combination of these. Execution of a program by at least one processor is not limited to execution of the entire program by a single processor, but may also include execution of the program by parallel processing in a multiprocessor constructed by multiple processors. In a multiprocessor, either symmetric or asymmetric multiprocessing may be performed. A multiprocessor may be tightly coupled or loosely coupled. A processor may be a multi-core processor.

[0020] The memory includes, but is not limited to, RAM, ROM, non-volatile RAM, PROM, EPROM, EEPROM, flash memory, magnetic data storage, optical data storage, registers, or a combination thereof. There may be one or more memories.

[0021] The output device is configured to output the CHL information. The output is performed visually or audibly. As shown in FIGS. 2 and 3 , the output device may be provided in the saddle-riding type vehicle, in equipment or a portable item of the rider, or in a user terminal. The output device itself may have a function of communicating with the outside via a network. The output device may be configured to be connected to a control device included in an apparatus in which the output device itself is provided (the saddle-riding type vehicle, the equipment or a portable item of the rider, or the user terminal), and to be controlled by the control device. In this case, the control device may have a function of communicating with the outside via a network. The output device outputs the CHL information visually or audibly. The output device is not particularly limited, and examples thereof include the following devices. - Meter (instrument panel) (presentation of visual output information) - HMD (Head-Mounted Display) (presentation of visual output information) - Speaker (presentation of auditory output information) - Intercom (headset) (presentation of auditory output information) - Mirror (presentation of visual output information by lights, etc.) - Road surface illumination light (presentation of visual output information) - Roadside display (presentation of visual output information) - HUD (Head-Up Display) combiner (presentation of visual output information) - HUD screen (presentation of visual output information) - Smartphone (presentation of visual and / or auditory output information) - Smartwatch (presentation of visual or auditory output information) - Smartglasses (presentation of visual or auditory output information) - Earphones (presentation of auditory output information) - Bone conduction earphones (presentation of auditory output information)

[0022] A saddle-type vehicle is a vehicle equipped with a saddle-type seat. A saddle-type vehicle is a vehicle configured so that a rider rides in a position straddling a saddle. In the present invention, a saddle-type vehicle is equipped with a handlebar and configured to be able to turn in a lean position. Saddle-type vehicles include motorcycles, three-wheeled motor vehicles having a pair of front or rear wheels, and four-wheeled motor vehicles having a pair of front and rear wheels, respectively. Motorcycles are not particularly limited, and examples include scooters, mopeds, off-road vehicles, and on-road vehicles. Saddle-type vehicles that can turn in a lean position have the property of making the rider feel a sense of unity and cooperation with the saddle-type vehicle, so the saddle-type vehicle-related CHL information output system according to the present invention is suitable for saddle-type vehicles that can turn in a lean position.

[0023] Rider equipment or personal items refer to devices that are worn or carried by the rider while riding a saddle-type vehicle and are capable of communication. The equipment or personal items are moved along with the rider by the saddle-type vehicle while the saddle-type vehicle is traveling. The rider here refers to a person riding in the saddle-type vehicle. Therefore, the rider includes not only the person driving the saddle-type vehicle but also a passenger. An example of a passenger is a tandem rider. Equipment is not particularly limited, and examples include HMDs, smart glasses, intercoms (headsets), shoes, boots, gloves, helmets, clothing, earphones, and bone conduction earphones. Personal items are not particularly limited, and examples include smartphones, electronic keys, personal digital assistants (PDAs), and smart watches. It is not necessary to strictly distinguish between equipment and personal items; items that fall under either or both categories are considered equipment or personal items.

[0024] A user terminal refers to a device that is not part of a rider's equipment or personal belongings and is used by a person who is not a rider (hereinafter also referred to as a user). However, a person may be a rider when riding a saddle-riding vehicle and a user when using the user terminal without riding a saddle-riding vehicle. The user terminal can constitute the system. Examples of users include, but are not limited to, a riding instructor for a saddle-riding vehicle or a riding school staff member. Examples of user terminals include, but are not limited to, a terminal carried by a riding instructor (e.g., a tablet or laptop PC) or a terminal installed in a riding school (e.g., a tablet, laptop PC, or desktop PC). When the system includes a user terminal, the riding instructor can obtain CHL information, which indicates a linguistic interpretation of the riding state of the saddle-riding vehicle, via the user terminal, thereby understanding the rider's riding skill. Since the information output from the user terminal is CHL information indicating the linguistic interpretation result, the riding instructor can quickly and easily grasp the rider's riding skills. After the rider has ridden the saddle-type vehicle, the riding instructor can explain the rider's riding skills while showing the user terminal on which the information is displayed. With this system, CHL information indicating the linguistic interpretation result can be quickly generated, and the information can be used for timely instruction to the rider and for discussions with the rider immediately after the ride.

[0025] The server is not a rider's equipment or a personal item, nor is it a user terminal. The server is not intended to be operated by either a rider or a user. However, the server may be operated by a person for purposes other than processing related to the system (e.g., maintenance).

[0026] A large-scale language model (LLM) refers to a language model with a large number of parameters. The large-scale language model is capable of executing natural language processing tasks. The large-scale language model may be a natural language generation model. The natural language generation model is based on the large-scale language model and is capable of creating a sentence consisting of new CHL information from input CHL information. Note that creating a sentence consisting of new CHL information from input CHL information is an example of a natural language processing task. The number of parameters in the large-scale language model may be, for example, 1 billion or more, 10 billion or more, or 100 billion or more. A language model refers to a model in which CHL is modeled using word occurrence probability. In one embodiment, the large-scale language model is a language model capable of performing inference using techniques such as zero-shot learning (ZSL), one-shot learning (OSL), or few-shot learning (FSL) without fine-tuning. In one embodiment, the large-scale language model is configured to execute a task and produce output in response to an input prompt. The prompt includes CHL information. The large-scale language model is not particularly limited, and examples thereof include GPT-3, GPT-4, GShard, Switch Transformer, Gopher, HyperCLOVA, etc. The large-scale language model is not included in the saddle-riding type vehicle-related CHL information output system, but can communicate with the saddle-riding type vehicle-related CHL information output system.

[0027] CHL information is linguistic information that can be understood, recognized, and memorized by humans. CHL information is based on the linguistic system used by humans in everyday conversation. CHL information is transmitted visually or audibly as text information or audio information.

[0028] The "at least one program" does not necessarily have to be stored in a single memory, but may be stored separately in different memories in devices that can communicate with each other. Execution of the "at least one program" performs data acquisition processing, prompt generation processing, linguistic interpretation processing, and output processing.

[0029] The data acquisition process is a process in which at least one processor acquires at least one type of saddle-riding-type vehicle-related data from at least one input device over time while the saddle-riding-type vehicle is traveling. The input device and the saddle-riding-type vehicle-related data are as described above. Acquisition over time refers to repeated acquisition of data over time. The specific manner of acquisition over time is not particularly limited, and may be, for example, periodic or aperiodic acquisition. Multiple types of data may be acquired over time. In this case, the acquisition timing of each piece of data may be the same or different.

[0030] The prompt generation process is a process in which at least one processor generates saddle-type vehicle-related prompt data using saddle-type vehicle-related data. In addition to the saddle-type vehicle-related data, other data may also be used. For example, the other data may include data obtainable via a network or data stored in a memory. The network is not particularly limited and may be constructed using various wireless networks. The communication method and communication protocol are not particularly limited. Examples of networks include the Internet, an in-vehicle network (e.g., CAN), an OBD-II network, and a mobile network such as LTE / 5G. The other data may include, for example, data used as a template for generating prompts and history data indicating past interactions. The other data may be customizable. For example, the other data may include rider biometric data obtained via an input device. The saddle-type vehicle-related prompt data can be linguistically interpreted by the LLM. The saddle-type vehicle-related prompt data has a smaller data volume than the saddle-type vehicle-related data. This means that the amount of generated saddle-riding-type vehicle-related prompt data is smaller than the amount of saddle-riding-type vehicle-related data used to generate it. Examples of saddle-riding-type vehicle-related prompt data include the following (i) to (iii).

[0031] (i) CHL information relating to the amount of change over time of a saddle-riding-type vehicle-related parameter indicated by the saddle-riding-type vehicle-related data and the method of evaluating that parameter. Because this prompt data is CHL information, it can be directly interpreted linguistically by the LLM. Furthermore, since it also includes data on the evaluation method, evaluation can be performed. Examples of the amount of change over time of the parameter are not particularly limited, and include, for example: - Amount of change over time of bank angle when turning - Amount of change over time of brake operation amount (operation pressure) when turning - Amount of change over time of running position when turning - Amount of change over time of vehicle speed

[0032] (ii) Visual data visually showing the amount of change over time in the saddle-riding-type vehicle-related parameters indicated by the saddle-riding-type vehicle-related data, and CHL information related to the evaluation method for the parameters. The visual data is, for example, data in the form of a graph showing the amount of change over time in the saddle-riding-type vehicle-related parameters. If the LLM is multimodal, the multimodal LLM can also interpret the visual data linguistically. Furthermore, since the evaluation method data is also included, evaluation can be performed. Specific examples are not particularly limited, and include, for example, the same examples as those in (i) above.

[0033] (iii) CHL information relating to a reading method for reading the saddle-riding-type vehicle-related data as the amount of change over time of the saddle-riding-type vehicle-related parameter and a method for evaluating the parameter. The saddle-riding-type vehicle-related data is numerical data. Therefore, the LLM cannot directly interpret the saddle-riding-type vehicle-related data linguistically. However, since the saddle-riding-type vehicle-related prompt data also includes data on the reading method, the LLM can read the saddle-riding-type vehicle-related data based on the reading method. Furthermore, since the saddle-riding-type vehicle-related prompt data also includes data on the evaluation method, the LLM can perform evaluation. Specific examples are not particularly limited, and include the same examples as in (i) above.

[0034] The saddle-riding-type vehicle-related prompt data is data for linguistically interpreting the saddle-riding-type vehicle-related data, and as shown in (i) to (iii) above, includes two elements: (I) how to read the saddle-riding-type vehicle-related data, and (II) how to evaluate the read data. Element (I) is, for example, how to obtain the characteristics or trends of the saddle-riding-type vehicle-related data. Element (II) is, for example, how to communicate the characteristics or trends to the rider. Element (II) may include, for example, advice to the rider. When the saddle-riding-type vehicle-related prompt data is generated using the saddle-riding-type vehicle-related data, elements (I) and (II) are included in the saddle-riding-type vehicle-related prompt data. Furthermore, another element (III) may be included. Examples of element (III) include, but are not limited to, a designation of the tone of the output (utterance), a designation of the creativity or accuracy used when generating the output (utterance), a designation of continuity or consistency with previous outputs (utterances), and a combination of at least two of these.

[0035] The linguistic interpretation process is a process in which at least one processor supplies saddle-riding vehicle-related prompt data to the LLM and acquires moving saddle-riding vehicle linguistic interpretation data from the LLM. The moving saddle-riding vehicle linguistic interpretation data is data generated by linguistic interpretation by the LLM.

[0036] The output process is a process of outputting CHL information indicating the linguistic interpretation result of the traveling state of the saddle riding type vehicle from at least one output device using the traveling saddle riding type vehicle linguistic interpretation data acquired from the LLM. The output process may be performed while the saddle riding type vehicle is traveling, or may be performed after the saddle riding type vehicle has been traveling. The output process may be performed so that the traveling saddle riding type vehicle linguistic interpretation data acquired from the LLM is output as is from the output device. The traveling saddle riding type vehicle linguistic interpretation data acquired from the LLM may be modified, and the output process may be performed using the modified traveling saddle riding type vehicle linguistic interpretation data. This output is performed visually and / or audibly.

[0037] The present invention can further employ the following configuration.

[0038] (2) The saddle-riding type vehicle related CHL information output system of (1), wherein the at least one program is further programmed to cause the at least one processor to execute the output process to output visual information consisting of images and / or video so as to at least partially overlap the CHL information in time with the CHL information.

[0039] According to the system of (2), CHL information and visual information are output so as to overlap in time. By using visual information in combination, the efficiency of information transmission to the rider can be improved. It is possible to increase the information transmitted to the rider and improve the impression given to the rider while suppressing the increase in output information. It is possible to achieve a higher level of interaction while suppressing the increase in load on hardware resources. The system of (2) can more effectively solve the above-mentioned unique problems.

[0040] The visual information may be stored in at least one memory in advance. The visual information may be received from outside the saddle-riding-type vehicle-related CHL information output system via a communication line such as the Internet. The visual information may be generated within the saddle-riding-type vehicle-related CHL information output system. The visual information may be generated by at least one processor. The visual information may be generated by the LLM. Images are static visual information. Videos are dynamic visual information. Images may consist of or include visualizations, or may include non-visualizations such as photo images. Videos may be animations or simulations, or may be live-action captures.

[0041] (3) The straddle-type vehicle related CHL information output system of (2), wherein the at least one program is further programmed to cause the at least one processor to execute a visual information generation process that generates the visual information using the at least one type of straddle-type vehicle related data acquired over time by the data acquisition process.

[0042] According to the system of (3), both the CHL information and the visual information are generated based on at least one type of saddle-ride type vehicle-related data and are output so as to overlap in time with each other. By increasing the correlation between the visual information and the CHL information, the efficiency of information transmission to the rider can be further improved. It is possible to increase the information transmitted to the rider and improve the impression given to the rider while suppressing an increase in the amount of output information. It is possible to achieve a higher level of interaction while suppressing an increase in the load on hardware resources. The system of (3) can more effectively solve the above-mentioned unique problems.

[0043] (4) The straddle-type vehicle related CHL information output system of (3), wherein the at least one program is further programmed to cause the at least one processor to execute the following in the visual information generation process: generate the visual information by the LLM using the at least one type of straddle-type vehicle related data acquired over time by the data acquisition process.

[0044] According to the system of (4), both the CHL information and the visual information are generated by the LLM based on at least one type of saddle-type vehicle-related data, and are output so as to overlap with each other in time. This makes it possible to improve the correlation between the visual information and the CHL information while reducing the load on hardware resources. The system of (4) can more effectively solve the above-mentioned unique problems.

[0045] (5) The saddle-riding type vehicle related CHL information output system according to any one of (2) to (4), wherein in the output process, the CHL information includes an explanation about or related to the visual information that is output so as to overlap in time with the CHL information.

[0046] According to the system of (5), the correlation between CHL information and visual information can be further enhanced. It is possible to have the rider recognize multiple types of highly relevant information in a timely manner. The efficiency of information transmission to the rider can be improved. It is possible to increase the information transmitted to the rider and improve the impression given to the rider while suppressing the increase in output information. It is possible to achieve a higher level of interaction while suppressing the increase in load on hardware resources. The system of (5) can more effectively solve the above-mentioned unique problems.

[0047] A description of or relating to visual information may be a description or explanation of the visual content contained in the visual information itself, or may be a description of how to view the visual content rather than the visual content contained in the visual information itself.

[0048] (6) The straddle-type vehicle-related CHL information output system according to any one of (2) to (5), wherein the at least one type of straddle-type vehicle-related data includes posture-related data and / or speed-related data, and running path data related to the running path of the straddle-type vehicle; the posture-related data includes vehicle posture data related to the posture of the straddle-type vehicle while running, and / or rider posture data related to the posture of a rider of the straddle-type vehicle while running; the speed-related data includes speed data related to the speed of the straddle-type vehicle, and / or acceleration data related to the acceleration of the straddle-type vehicle; and the visual information is configured by associating at least one type of information among the posture of the straddle-type vehicle, the posture of the rider, the vehicle speed of the straddle-type vehicle, and the acceleration of the straddle-type vehicle with the running path.

[0049] According to the system (6), both the CHL information and the visual information are related to the rider's driving technique. As described above, saddle-type vehicles are vehicles that tend to be enjoyed for the act of driving itself, so information related to driving technique is interesting to the rider and is information that makes the rider feel a sense of unity and cooperation with the saddle-type vehicle. The system (6) can provide the rider with information related to driving technique in a timely manner while reducing the load on hardware resources, thereby improving the impression given to the rider. A higher level of interaction can be achieved while reducing the load on hardware resources.

[0050] The vehicle attitude data is generated using, for example, motion capture. The motion capture is not particularly limited as long as it can digitize the movements of a person or an object and input them into a computer. The vehicle attitude data may be generated using motion capture and an IMU or GNSS receiving unit mounted on the saddle-type vehicle. The vehicle attitude data relates, for example, to the inclination of the vehicle body or the steering angle during a turn. The rider attitude data is generated, for example, using motion capture. The rider attitude data may be generated based on signals from inertial sensors such as IMUs attached to various parts of the rider or sensors attached to the joints of the rider that detect angles or displacements. The rider attitude data may be data generated by analyzing images of a person captured by a camera. The rider attitude data relates, for example, to the rider's attitude during a turn. The traveling trajectory data is generated, for example, using GNSS and sensors included in the saddle-type vehicle. The sensor possessed by the saddle-riding vehicle may be, for example, an IMU, a sensor that detects the steering angle of the steering wheels or steering skis, or a sensor that contributes to detecting the speed of the saddle-riding vehicle in the forward or traveling direction. The travel trajectory data may be data generated without using GNSS. For example, the travel trajectory data may be data generated using a radio beacon. In this case, the saddle-riding vehicle is equipped with a receiver capable of receiving electromagnetic waves such as radio waves transmitted from a radio station. The travel trajectory data may be generated based on data generated based on radio waves received by the receiver. The travel trajectory data may be generated based on map data and data generated based on radio waves received by the receiver. The travel trajectory data is output, for example, as an image showing the travel trajectory. The image showing the travel trajectory is output, for example, superimposed on a map image.

[0051] (7) The saddle-riding vehicle-related CHL information output system of (6), wherein the traveling trajectory includes a J-shaped trajectory related to a linear trajectory of the saddle-riding vehicle and a circular arc trajectory continuing therefrom, and the visual information is configured by associating at least one type of information with the J-shaped trajectory when the saddle-riding vehicle is traveling on the J-shaped trajectory: the posture of the saddle-riding vehicle, the posture of the rider, the vehicle speed of the saddle-riding vehicle, and the acceleration of the saddle-riding vehicle.

[0052] When a saddle-type vehicle travels in a J-shaped trajectory, the rider's driving technique tends to be more easily reflected in the saddle-type vehicle-related data. According to the system of (7), the information provided by the system is more interesting to the rider, and the rider is more likely to feel a sense of unity and cooperation with the saddle-type vehicle. The system of (7) can provide the rider with information related to driving technique in a timely manner and in a more interesting format while reducing the load on hardware resources, thereby improving the impression given to the rider. A higher level of interaction can be achieved while reducing the load on hardware resources.

[0053] The J-shaped trajectory refers to a linear trajectory and a circular arc trajectory continuing from the linear trajectory (see FIGS. 6( a) and 6(b)). The linear trajectory reflects the rider's operation before the turn, i.e., the rider's driving technique when approaching the turn. The circular arc trajectory reflects the rider's operation during the turn, in other words, the rider's driving technique. When transitioning from a state in which the vehicle is traveling along a linear trajectory to a state in which the vehicle is traveling along an arc trajectory, the rider's operations switch with respect to weight shift, steering, and throttle and brake control. Because the J-shaped trajectory includes such a change in operation and the trajectory before and after that, the rider's driving technique tends to be easily reflected in the saddle-riding vehicle-related data. The linear trajectory does not necessarily have to be a strictly straight line. The circular arc trajectory does not necessarily have to be a perfect circular arc.

[0054] (8) The saddle-riding-type vehicle-related CHL information output system of any one of (1) to (7), wherein the at least one program is further programmed to cause the at least one processor to execute the following: a comparison data acquisition process for acquiring comparison saddle-riding-type vehicle-related data to be used for comparison with the at least one type of saddle-riding-type vehicle-related data; in the prompt generation process, using the at least one type of saddle-riding-type vehicle-related data and the comparison saddle-riding-type vehicle-related data, generate the saddle-riding-type vehicle-related prompt data so as to include a request to the LLM to cause the LLM to compare the at least one type of saddle-riding-type vehicle-related data with the comparison saddle-riding-type vehicle-related data; in the linguistic interpretation process, obtain from the LLM the linguistic interpretation data for the moving saddle-riding type vehicle including data related to the comparison result; and in the output process, output the CHL information linguistically indicating a comment on the driving state of the saddle-riding type vehicle based on the comparison result.

[0055] In the system of (8), a comment about the riding condition of the saddle-riding type vehicle is linguistically output based on at least the results of comparing the saddle-riding type vehicle-related data with the comparative saddle-riding type vehicle-related data. The comparative saddle-riding type vehicle-related data is, for example, saddle-riding type vehicle-related data related to the riding condition of another saddle-riding type vehicle, and in that case, the comment relates to the results of the comparison with the other saddle-riding type vehicle. Such comparative information is more interesting to the rider and makes it easier for the rider to feel a sense of unity and cooperation with the saddle-riding type vehicle. The system of (8) can provide the rider with information related to driving technique in a timely manner and in a more interesting format while reducing the load on hardware resources, thereby further improving the rider's impression. A higher level of interaction can be achieved while reducing the load on hardware resources.

[0056] The comparative straddle-type vehicle-related data is not limited to the above examples, and may include the following: (i) Data obtained from one straddle-type vehicle that has actually traveled the same or a similar route, or data obtained by processing such data. (ii) Data obtained by processing data obtained from multiple straddle-type vehicles that have actually traveled the same or a similar route (e.g., average data). (iii) Data obtained from a simulation of a straddle-type vehicle traveling the same or a similar route. (iv) Data obtained by extracting multiple feature points from any one of (i) to (iii) above so as to be comparable with at least one type of straddle-type vehicle-related data. Unlike (i) and (ii) above, the data (iii) above is not data obtained from an actual travel of a straddle-type vehicle. The data (iv) above has a smaller amount of data than any one of (i) to (iii) above from which it was originally derived. The comparative straddle-type vehicle-related data may also be stored in advance in at least one memory. The comparison straddle-type vehicle-related data may be received from outside the straddle-type vehicle-related CHL information output system via a communication line such as the Internet. The comparison straddle-type vehicle-related data may be generated within the straddle-type vehicle-related CHL information output system. The comparison straddle-type vehicle-related data may be generated by at least one processor. The comparison straddle-type vehicle-related data may be generated by the LLM.

[0057] (9) The saddle-riding type vehicle-related CHL information output system of (8), wherein the comparison is performed by comparing a plurality of feature points included in the comparison saddle-riding type vehicle-related data with the at least one type of saddle-riding type vehicle-related data, and the comments include high-similarity comments about feature points that are highly similar to each other, and / or high-difference comments about feature points that are highly different from each other.

[0058] In the system (9), high similarity comments and / or high dissimilarity comments are output. These comments tend to more clearly express the results of the comparison. The system (9) can provide comments that are more interesting to the rider and that make the rider feel more connected and collaborative with the saddle-type vehicle, while suppressing an increase in the number of comments. This can further improve the impression given to the rider while suppressing an increase in the number of comments. A higher level of interaction can be achieved while suppressing the load on hardware resources.

[0059] The present invention can further employ the following configuration: (10) The saddle-riding-type vehicle-related CHL information output system according to any one of (1) to (9), including the at least one input device, the at least one output device, and a saddle-riding-type vehicle-related IPG that can communicate with each of the at least one input device, the at least one output device, and the LLM via a network, and that has the at least one memory and the at least one processor.

[0060] The system (10) includes a saddle-type vehicle-related IPG, so customization by entities such as saddle-type vehicle manufacturers can be easily realized.

[0061] The present invention can further employ the following configurations: (11) A straddle-type vehicle-related IPG included in the straddle-type vehicle-related CHL information output system of (10).

[0062] The saddle-type vehicle-related IPG (11) can be easily customized by entities such as saddle-type vehicle manufacturers.

[0063] According to the present invention, by effectively and efficiently enhancing the sense of cooperation between the saddle-type vehicle and the rider, it is possible to build a system that can improve the rider's well-being and reduce the load on the hardware resources that make up the system.

[0064] FIG. 1( a) is a system overview diagram illustrating a straddle-type vehicle-related CHL information output system according to an embodiment of the present invention, and FIG. 1( b) is a flowchart illustrating processing performed by the straddle-type vehicle-related CHL information output system. FIGS. 2( a) to 2( x) are diagrams illustrating devices in which an input device (I), a processor (P), a memory (M), and an output device (O) are provided in the straddle-type vehicle-related CHL information output system according to the present invention. FIGS. 3( a) to 3( x) are tables illustrating devices in which an input device, a processor, and an output device are provided in the straddle-type vehicle-related CHL information output system according to the present invention. FIG. 4 is a flowchart illustrating processing performed by a straddle-type vehicle-related CHL information output system according to a first modified embodiment of the present invention. FIGS. 5( a) and 5( b) are diagrams illustrating a J-shaped trajectory included in visual information output by the straddle-type vehicle-related CHL information output system according to the first modified embodiment. FIG. 6 is a flowchart illustrating processing performed by a straddle-type vehicle-related CHL information output system according to a second modified embodiment of the present invention.

[0065] FIG. 1A is a system overview diagram for explaining a straddle-type vehicle-related CHL information output system 1 according to an embodiment of the present invention.

[0066] The straddle-type vehicle-related CHL information output system 11 (hereinafter simply referred to as the system 11) includes at least one input device 13, at least one output device 14, and a straddle-type vehicle-related interactive prompt generator 1 (hereinafter also referred to as the straddle-type vehicle-related IPG 1). The at least one input device 13 is provided in at least one of a straddle-type vehicle 16 and a rider's equipment 17 and / or a rider's portable item 17. The at least one processor 2 acquires straddle-type vehicle-related data from the at least one input device 13. Examples of the input device 13 shown in the figure include a GPS module, an IMU, and an ECU. As described above, the ECU does not function as an input device alone, but functions as an input device in combination with the GPS module or the IMU. The at least one processor 2 acquires straddle-type vehicle-related data from each of the GPS module and the IMU via the ECU. Of course, the input device 13 is not limited to this example.

[0067] The saddle-ride type vehicle-related IPG 1 includes at least one processor 2, at least one memory 3, and at least one communication module 4. The memory 3 stores programs for executing (A) data acquisition processing, (B) prompt generation processing, (C) linguistic interpretation processing, and (D) output processing. The at least one processor 2 executes the programs. The number of processors 2, memories 3, and communication modules 4 is not particularly limited, and each may be one or more.

[0068] The hardware configuration of the system 11 is not particularly limited. The output device 14 may also have the functions of the saddle-riding-type vehicle-related IPG 1. If the saddle-riding-type vehicle-related IPG 1 is separate from the output device 14, the saddle-riding-type vehicle-related IPG 1 may be configured by a single server 19 or multiple servers 19 connected to each other so that they can communicate with each other. In this case, the multiple servers 19 may be configured to provide cloud computing services. The communication module 4 enables communication between the processor 2 and each of the LLM (Large Scale Language Model) 6, the input device 13, and the output device 14.

[0069] Each of the multiple output devices 14 can communicate with the saddle-riding vehicle-related IPG 1 via the network 12. The number of output devices 14 is not particularly limited. At least one output device is required. The output device 14 may be provided in the saddle-riding vehicle 16, in the rider's equipment 17 or personal belongings 17, or in the user terminal 18. Examples of the output device 14 shown in the figure include a meter on the saddle-riding vehicle 16, a helmet HUD, and a smartphone. Of course, the output device 14 is not limited to these examples.

[0070] The network 12 enables communication between the multiple output devices 14, the saddle-ride type vehicle-related IPG 1, and the LLM 6. The network 12 is not particularly limited and can be constructed using various wireless networks. There are no particular limitations on the communication method or communication protocol.

[0071] The LLM 6 is stored in one server or in multiple servers that are communicatively connected to each other. The LLM 6 can communicate with the system 11. This is the same as the LLM 6 being able to communicate with the saddle-ride type vehicle-related IPG 1 in this embodiment. The LLM 6 outputs to the system 11 in response to input from the system 11. Communication between the LLM 6 and the system 11 is performed using at least CHL information.

[0072] Fig. 1(b) is a flowchart for explaining the processing performed by the output system 11. In this embodiment, the saddle-riding-type vehicle-related data is data related to the bank angle of the saddle-riding-type vehicle 16 generated by the IMU and data related to the traveling position of the saddle-riding-type vehicle 16 generated by the GPS module. Note that the processing shown in Fig. 1(b) can be implemented in any of the system configurations shown in Figs. 2 and 3(a) to 3(x).

[0073] [Data Acquisition Process (A)] At least one processor 2 acquires at least one type of saddle-riding-type vehicle-related data over time from an input device 13 provided on the saddle-riding-type vehicle 16 or on the rider's equipment 17 or portable equipment 17 while the saddle-riding-type vehicle 16 is traveling (step S11). Specifically, the at least one processor 2 acquires data related to the bank angle of the saddle-riding-type vehicle 16 over time from the IMU, and acquires data related to the traveling position of the saddle-riding-type vehicle 16 over time from the GPS module.

[0074] [Prompt Generation Process (B)] At least one processor 2 generates saddle-riding-type vehicle-related prompt data using at least one type of saddle-riding-type vehicle-related data acquired in step S11 (step S12). In this embodiment, the saddle-riding-type vehicle-related prompt data is data indicating the following CHL information: - Amount of change over time in the bank angle of the saddle-riding-type vehicle when turning, - A change over time in the running position of the saddle-riding-type vehicle when turning, and - A method for evaluating both of the above parameters.

[0075] [Linguistic Interpretation Process (C)] At least one processor 2 supplies the saddle-ride type vehicle-related prompt data generated in step S12 to the LLM 6 (step S13).

[0076] The LLM 6 linguistically interprets the saddle-riding-type vehicle-related prompt data supplied in step S13 (step S21). In this embodiment, the LLM 6 uses the supplied saddle-riding-type vehicle-related prompt data to evaluate the relationship between the traveling path and bank angle during a turn based on the evaluation method. For example, if the relationship between the traveling path and bank angle during a turn matches the evaluation method, the LLM 6 generates traveling saddle-riding-type vehicle linguistic interpretation data such as, "The bank angle during the turn was ideal."

[0077] Next, the at least one processor 2 acquires the traveling saddle-riding vehicle linguistic interpretation data generated in step S21 from the LLM 6 (step S14).

[0078] [Output Process (D)] At least one processor 2 supplies the traveling saddle-riding type vehicle linguistic interpretation data acquired from the LLM 6 in step S14 to the output device 14 (step S15).

[0079] The output device 14 outputs CHL information indicating the linguistic interpretation result regarding the running state of the saddle riding type vehicle based on the running saddle riding type vehicle linguistic interpretation data supplied from at least one processor 2 (step S31). In this embodiment, the CHL information such as "The bank angle during cornering was ideal" is displayed on the meter of the saddle riding type vehicle 16.

[0080] As described above, in this embodiment, the state of a saddle-riding vehicle 16 capable of turning in a lean position is grasped using IMU and GPS data, i.e., multiple types of data, and an evaluation of the rider's driving skill is provided. Here, a saddle-riding vehicle 16 capable of turning in a lean position has a greater variety of information available regarding driving operations and running conditions than a non-lean vehicle. Multiple types of data can be used in combination. This makes it easier to grasp the state of a saddle-riding vehicle 16 capable of turning in a lean position. In other words, a saddle-riding vehicle 16 capable of turning in a lean position has a greater variety of information available than a non-lean vehicle, so using multiple types of data in combination to grasp the state of the saddle-riding vehicle 16 is suitable for a saddle-riding vehicle 16 capable of turning in a lean position. Note that the data used to evaluate the rider's driving skill is not limited to the above data.

[0081] (Variation 1) A straddle-type vehicle-related CHL information output system according to Variation 1 of the embodiment of the present invention will be described with reference to FIG. 4 . In Variation 1, the memory 3 stores programs for executing (A) data acquisition processing, (B) prompt generation processing, (C) linguistic interpretation processing, (D1) output processing, and (E) visual information generation processing. At least one processor 2 executes the programs. In Variation 1, the (A) data acquisition processing, (B) prompt generation processing, and (C) linguistic interpretation processing are the same as those in the above embodiment. Therefore, a description of these will be omitted. The (D1) output processing and (E) visual information generation processing will be described below.

[0082] [Visual Information Generation Process (E)] At least one processor 2 generates visual information using at least one type of saddle-riding-type vehicle-related data acquired in step S11. In Modification 1, at least one processor 2 generates visual information using the at least one type of saddle-riding-type vehicle-related data acquired in step S11, via the LLM 6 (step S22). More specifically, the process is as follows: At least one processor 2 supplies the saddle-riding-type vehicle-related data acquired in step S11 to the LLM 6 (step S16). The LLM 6 generates visual information using the saddle-riding-type vehicle-related data provided in step S16 (step S22). At least one processor 2 acquires the visual information generated in step S22 from the LLM 6 (step S17). Note that in the example shown in FIG. 4 , the visual information generation process (E) is performed after the linguistic interpretation process (C), but the visual information generation process (E) may be performed before the linguistic interpretation process (C) or before the prompt generation process (B).

[0083] In the first modification, at least one type of saddle-riding-type vehicle-related data includes posture-related data and / or speed-related data, and travel path data related to the travel path of the saddle-riding-type vehicle 16. As shown in FIG. 5A , the posture-related data includes vehicle posture data related to the posture of the saddle-riding-type vehicle 16 while traveling, and / or rider posture data related to the posture of the rider of the saddle-riding-type vehicle 16 while traveling. The speed-related data includes speed data related to the speed of the saddle-riding-type vehicle 16 and / or acceleration data related to the acceleration of the saddle-riding-type vehicle 16. The visual information is made up of images or videos. The visual information is configured by associating at least one type of information from the posture of the saddle-riding-type vehicle 16, the posture of the rider, the vehicle speed of the saddle-riding-type vehicle 16, and the acceleration of the saddle-riding-type vehicle 16 with the travel path. Note that, although the LLM 6 generates the visual information in the example shown in FIG. 4 , the visual information may be stored in advance in at least one memory 3, or may be received from outside the system 11 via a communication line such as the Internet. In Modification 1, as shown in Figure 5(a), the traveling trajectory includes a J-shaped trajectory related to a linear trajectory of the saddle riding type vehicle 16 and a circular arc trajectory continuing therefrom. Note that the traveling trajectory is not limited to a J-shaped trajectory. In Modification 1, as shown in Figure 5(b), the visual information is configured by associating at least one type of information with the J-shaped trajectory when the saddle riding type vehicle 16 is traveling along the J-shaped trajectory, from the posture of the saddle riding type vehicle 16, the posture of the rider, the vehicle speed of the saddle riding type vehicle 16, and the acceleration of the saddle riding type vehicle 16.

[0084] [Output Process (D1)] At least one processor 2 supplies the traveling saddle-riding vehicle linguistic interpretation data acquired from the LLM 6 in step S14 and the visual information acquired from the LLM 6 in step S17 to the output device 14 (step S15A). At least one processor 2 outputs the visual information in step S31A so as to at least partially overlap with the CHL information in time. Note that in the example shown in FIG. 4 , the output of the CHL information and the visual information starts at the same time. However, for example, the output of the CHL information may start before the visual information, or the output of the CHL information may start after the visual information. The same applies to the end of the output of the CHL information and the visual information. In Modification 1, the CHL information may include an explanation about or related to the visual information that is output so as to overlap in time with the CHL information.

[0085] (Variation 2) A straddle-type vehicle-related CHL information output system according to Variation 2 of the embodiment of the present invention will be described with reference to Figure 6. In Variation 2, the memory 3 stores programs for executing (A) data acquisition processing, (B1) prompt generation processing, (C1) linguistic interpretation processing, (D2) output processing, and (F) comparison data acquisition processing. At least one processor 2 executes these programs. In Variation 2, the (A) data acquisition processing is the same as in the above embodiment, and therefore its description will be omitted. The (B1) prompt generation processing, (C1) linguistic interpretation processing, (D2) output processing, and (F) comparison data acquisition processing will be described below.

[0086] [Comparison Data Acquisition Process (F)] At least one processor 2 acquires comparison saddle-riding-type vehicle-related data to be used for comparison with at least one type of saddle-riding-type vehicle-related data (step S18). In Modification 2, the comparison saddle-riding-type vehicle-related data is acquired from outside the system 11 via a communication line such as the Internet. The comparison saddle-riding-type vehicle-related data may be stored in advance in at least one memory, or may be generated within the system 11. The comparison data acquisition process (F) only needs to be performed before the prompt generation process (B1). The comparison saddle-riding-type vehicle-related data is, for example, saddle-riding-type vehicle-related data relating to the traveling state of another saddle-riding type vehicle.

[0087] [Prompt Generation Process (B1)] In step S12A, at least one processor 2 uses the at least one type of saddle-riding vehicle-related data acquired in step S11 and the comparison saddle-riding vehicle-related data acquired in step S18 to generate saddle-riding vehicle-related prompt data that includes a request to the LLM to cause the LLM to compare the at least one type of saddle-riding vehicle-related data with the comparison saddle-riding vehicle-related data.

[0088] [Linguistic Interpretation Process (C1)] At least one processor 2 supplies the saddle-riding-type vehicle-related prompt data generated in step S12A to the LLM 6 (step S13A). The LLM 6 linguistically interprets the saddle-riding-type vehicle-related prompt data supplied in step S13A (step S21A). The LLM 6 generates moving saddle-riding-type vehicle linguistic interpretation data including data related to the comparison results. The at least one processor 2 acquires the moving saddle-riding-type vehicle linguistic interpretation data including data related to the comparison results generated in step S21A from the LLM 6 (step S14A). [Output Process (D2)] At least one processor 2 supplies the moving saddle-riding-type vehicle linguistic interpretation data acquired from the LLM 6 in step S14A to the output device 14 (step S15B). In step S31A, the at least one processor 2 outputs CHL information linguistically indicating a comment on the driving state of the saddle-riding-type vehicle 16 based on the comparison results. In the second modification, the comparison is performed by comparing a plurality of feature points included in the comparison saddle-riding type vehicle-related data with a plurality of feature points included in at least one type of saddle-riding type vehicle-related data. In the second modification, the comments include high-similarity comments about feature points that are highly similar to each other, and / or high-dissimilarity comments about feature points that are highly different from each other.

[0089] In this embodiment, an evaluation of a rider's driving skill is provided using data from an IMU and a GPS. However, the present invention is not limited to the above embodiment. For example, information on road conditions may be provided using data from an IMU and a GPS. In this case, weather information obtained from the Internet may also be used. The data used to provide information on road conditions is not limited to the above data. Road conditions are an example of the environment surrounding a saddle-riding vehicle. In a saddle-riding vehicle that can turn in a lean position, the rider's weight shifts, resulting in the vehicle leaning. Weight shifting is an operation that does not occur in non-lean vehicles. Manual two-wheeled vehicles require separate operations with both hands and feet. Riders of all types of motorcycles, not just manual vehicles, tend to input a greater variety of operations while driving. Therefore, compared to non-lean vehicles, riders want to be aware of the surrounding environment. The surrounding environment can be understood from the state of the saddle-riding vehicle. As described above, the state of a saddle-riding vehicle can be easily understood by using multiple types of data. In other words, since a saddle-type vehicle can obtain more information than a non-lean vehicle, using a combination of multiple types of data to understand the surrounding environment is suitable for a saddle-type vehicle that can turn in a lean position. The present invention can be implemented in other embodiments, and various modifications can be added.

[0090] 1: Saddle-ride type vehicle related IPG (saddle-ride type vehicle related interactive prompt generator) 2: Processor 3: Memory 4: Communication module 6: LLM 11: Saddle-ride type vehicle related CHL information output system (system) 12: Network 13: Input device 14: Output device

Claims

1. A saddle-riding-type vehicle-related CHL (Communicative Human Language) information output system, the saddle-riding-type vehicle-related CHL information output system comprising: at least one memory; at least one processor capable of communicating with an LLM (Large Scale Language Model), connected to the memory, and configured to execute at least one program stored in the memory; at least one input device provided on the saddle-riding-type vehicle or on equipment or carried by a rider of the saddle-riding-type vehicle, configured to acquire or generate at least one type of saddle-riding-type vehicle-related data that is related to the traveling state of the saddle-riding-type vehicle and changes as the saddle-riding-type vehicle is traveling, and to supply the at least one type of saddle-riding-type vehicle-related data to the at least one processor; and at least one output device configured to output CHL information, the at least one program comprising: a data acquisition process for acquiring the at least one type of saddle-riding-type vehicle-related data from the at least one input device over time while the saddle-riding-type vehicle is traveling; a prompt generation process that generates saddle-riding-type vehicle-related prompt data that can be linguistically interpreted by the LLM and has a reduced amount of data, using the at least one type of saddle-riding-type vehicle-related data acquired over time by the data acquisition process; a linguistic interpretation process that supplies the saddle-riding-type vehicle-related prompt data to the LLM and acquires from the LLM linguistically interpreted data for the saddle-riding-type vehicle that has been generated by the linguistic interpretation by the LLM; and an output process that outputs CHL information indicating the linguistic interpretation result for the driving state of the saddle-riding type vehicle from the at least one output device, using the linguistically interpreted data for the saddle-riding type vehicle acquired from the LLM.

2. A straddle-type vehicle related CHL information output system as set forth in claim 1, wherein said at least one program is further programmed to cause said at least one processor to execute the output process of outputting visual information consisting of images and / or video so as to at least partially overlap in time with said CHL information.

3. A straddle-type vehicle related CHL information output system according to claim 2, wherein said at least one program is further programmed to cause said at least one processor to execute a visual information generation process for generating said visual information using said at least one type of straddle-type vehicle related data acquired over time by said data acquisition process.

4. A straddle-type vehicle related CHL information output system according to claim 3, wherein said at least one program is further programmed to cause said at least one processor to execute the following in said visual information generation process: generate said visual information by said LLM using said at least one type of straddle-type vehicle related data acquired over time by said data acquisition process.

5. A straddle-type vehicle related CHL information output system according to any one of claims 2 to 4, wherein in the output process, the CHL information includes an explanation about or related to the visual information that is output so as to overlap in time with the CHL information.

6. A straddle-type vehicle-related CHL information output system according to any one of claims 2 to 5, wherein the at least one type of straddle-type vehicle-related data includes posture-related data and / or speed-related data, and running path data relating to the running path of the straddle-type vehicle, the posture-related data includes vehicle posture data relating to the posture of the straddle-type vehicle while running, and / or rider posture data relating to the posture of the rider of the straddle-type vehicle while running, the speed-related data includes speed data relating to the speed of the straddle-type vehicle, and / or acceleration data relating to the acceleration of the straddle-type vehicle, and the visual information is configured by associating at least one type of information from the posture of the straddle-type vehicle, the posture of the rider, the vehicle speed of the straddle-type vehicle, and the acceleration of the straddle-type vehicle with the running path.

7. A straddle-type vehicle related CHL information output system according to claim 6, wherein the travel trajectory includes a J-shaped trajectory related to a linear trajectory of the straddle-type vehicle and a circular arc trajectory continuing therefrom, and the visual information is configured by associating at least one type of information with the J-shaped trajectory when the straddle-type vehicle is traveling along the J-shaped trajectory: the posture of the straddle-type vehicle, the posture of the rider, the vehicle speed of the straddle-type vehicle, and the acceleration of the straddle-type vehicle.

8. A straddle-type vehicle-related CHL information output system according to any one of claims 1 to 7, wherein the at least one program is further programmed to cause the at least one processor to execute the following: a comparison data acquisition process for acquiring comparison straddle-type vehicle-related data to be used for comparison with the at least one type of straddle-type vehicle-related data; in the prompt generation process, using the at least one type of straddle-type vehicle-related data and the comparison straddle-type vehicle-related data, generate the straddle-type vehicle-related prompt data so as to include a request to the LLM to cause the LLM to compare the at least one type of straddle-type vehicle-related data with the comparison straddle-type vehicle-related data; in the linguistic interpretation process, obtain from the LLM the linguistic interpretation data for the moving straddle-type vehicle including data related to the results of the comparison; and in the output process, output the CHL information linguistically indicating a comment on the driving state of the straddle-type vehicle based on the results of the comparison.

9. A straddle-type vehicle-related CHL information output system according to claim 8, wherein the comparison is performed by comparing a plurality of feature points contained in the comparison straddle-type vehicle-related data and the at least one type of straddle-type vehicle-related data, and the comments include high-similarity comments about feature points that are highly similar to each other, and / or high-difference comments about feature points that are highly different from each other.

10. A straddle-type vehicle-related CHL information output system according to any one of claims 1 to 9, comprising: the at least one input device; the at least one output device; and a straddle-type vehicle-related IPG (Interactive Prompt Generator) capable of communicating with each of the at least one input device, the at least one output device, and the LLM via a network, and having the at least one memory and the at least one processor.

11. A straddle-type vehicle-related IPG included in the straddle-type vehicle-related CHL information output system according to claim 10.

Images