Strategy proposal system, instruction generation system, strategy proposal method, and strategy proposal program

Abstract

A strategy proposal system 100 that uses a generative AI model 92 to propose a running strategy in a running race includes: an acquisition unit 64 that acquires target athlete information including information on past running of a target athlete; an instruction generation unit 74 that generates a strategy generation instruction for causing the generative AI model 92 to output running strategy information corresponding to a virtual competitor in a predetermined race on the basis of the target athlete information; a strategy generation unit 76 that generates running strategy information corresponding to the strategy generation instruction by inputting the strategy generation instruction into the generative AI model 92; and an output unit 78 that outputs the running strategy information.

Description

Strategy proposal system, instruction generation system, strategy proposal method, and strategy proposal program

[0001] The present disclosure relates to a technique for proposing a running strategy in a running race.

[0002] In a running race such as a marathon, in order to maximize the performance of a runner aiming for victory or a high-ranking prize, it is important to formulate an appropriate running strategy. Conventionally, such a running strategy has mainly been formulated through communication with the runner based on the runner's characteristics and past performance data by a coach.

[0003] Patent Document 1 describes a technique for calculating a recommended pace in a long-distance race such as a marathon, in which a recommended running pace is calculated from the running pace actually run and acquired by the judged runner.

[0004] Japanese Patent No. 6963112

[0005] In order to aim for victory or a high-ranking prize in a running race, it is important to formulate a running strategy considering the characteristics of competing opponents. However, the technique described in Patent Document 1 does not consider competing opponents.

[0006] The present disclosure has been made in view of such problems, and an object thereof is to provide a technique capable of proposing a running strategy for winning against competing opponents in a running race.

[0007] To solve the above problems, a strategy proposal system according to an aspect of the present disclosure is a strategy proposal system that proposes a running strategy in a running race using a generation AI model, where the generation AI model is a model that generates and outputs information corresponding to a generation instruction when the generation instruction is input, an acquisition unit that acquires target runner information including information on the past running of a target runner, an instruction generation unit that generates a strategy generation instruction for outputting running strategy information corresponding to a virtual competing opponent in a predetermined race to the generation AI model based on the target runner information, a strategy generation unit that generates running strategy information corresponding to the strategy generation instruction by inputting the strategy generation instruction to the generation AI model, and an output unit that outputs the running strategy information.

[0008] Another aspect of the present disclosure is an instruction generation system. This system is an instruction generation system that generates instructions to an AI model for proposing running strategies in a running race using the AI ​​model, wherein the AI ​​model generates and outputs information corresponding to the generation instructions when the generation instructions are input, and comprises an acquisition unit that acquires target athlete information including information on the target athlete's past running, and an instruction generation unit that generates strategy generation instructions to cause the AI ​​model to output running strategy information corresponding to virtual competitors in a predetermined race based on the target athlete information.

[0009] Another aspect of this disclosure is a strategy proposal method. This method is a strategy proposal method for proposing a running strategy in a running race using a generative AI model, wherein the generative AI model is a model that generates and outputs information corresponding to a generation instruction when a generation instruction is input, and includes the steps of: a computer acquiring target athlete information including information on the target athlete's past running; a computer generating a strategy generation instruction based on the target athlete information to cause the generative AI model to output running strategy information corresponding to a virtual competitor in a predetermined race; a computer generating running strategy information corresponding to the strategy generation instruction by inputting the strategy generation instruction into the generative AI model; and a computer outputting the running strategy information.

[0010] Another aspect of this disclosure is a strategy proposal program. This program is a strategy proposal program that proposes running strategies in a running race using a generation AI model, wherein the generation AI model generates and outputs information corresponding to a generation instruction when a generation instruction is input, and enables a computer to implement the following functions: acquiring target athlete information including information on the target athlete's past running; generating strategy generation instructions based on the target athlete information to cause the generation AI model to output running strategy information corresponding to a virtual competitor in a predetermined race; inputting the strategy generation instructions into the generation AI model to generate running strategy information corresponding to the strategy generation instructions; and outputting the running strategy information.

[0011] Furthermore, any combination of the above components, or any substitution of the components or expressions of this disclosure between methods, apparatus, programs, temporary or non-temporary storage media storing programs, systems, etc., are also valid forms of this disclosure.

[0012] According to this disclosure, it is possible to propose a running strategy for beating competitors in a running race.

[0013] This figure shows the configuration of the strategy proposal system according to the first embodiment. This is a functional block diagram showing each component of the strategy proposal system in Figure 1. This is a functional block diagram showing each function of the strategy proposal server in Figure 1. This is a flowchart that schematically shows the strategy proposal processing process in the strategy proposal server in Figure 1. This figure shows an example of a screen that displays and outputs target player information. This figure shows an example of a screen that displays and outputs target player characteristics. This figure shows a first example of race course information included in race information. This figure shows a first example of a screen that displays and outputs running strategy information. This figure shows a second example of race course information included in race information. This figure shows an example of virtual competitor information included in race information. This figure shows a second example of a screen that displays and outputs running strategy information. This figure shows a third example of a screen that displays and outputs running strategy information. This is a functional block diagram showing each function of the instruction generation server included in the instruction generation system according to the second embodiment.

[0014] The present disclosure will be described below with reference to the drawings, based on preferred embodiments. In embodiments and modifications, the same or equivalent components will be denoted by the same reference numerals, and redundant descriptions will be omitted as appropriate.

[0015] [First Embodiment] Figure 1 shows the configuration of the strategy proposal system 100 according to the first embodiment. The strategy proposal system 100 is a system that proposes a running strategy for a target athlete 10 in a running race such as a marathon (hereinafter also simply referred to as "race") using a generative AI model. Hereinafter, the target athlete 10 refers to an athlete who is the target of strategy proposals by the strategy proposal system 100 of this disclosure. The strategy proposal system 100 is envisioned to be used, for example, during a strategy meeting before a race. The strategy meeting is held, for example, from one month before the race until just before, especially a few days before, and the participants of the strategy meeting include the target athlete 10 and a coach.

[0016] First, let's explain the overview of the strategy proposal system 100. The strategy proposal system 100 comprises an information terminal 50 and a strategy proposal server 60 that can send and receive data from each other. The information terminal 50 transmits target athlete information, including information about the target athlete 10's past races, to the strategy proposal server 60. The strategy proposal server 60 generates a strategy generation instruction based on the received target athlete information. The strategy generation instruction is an instruction to cause the generating AI model to output racing strategy information corresponding to virtual competitors in a predetermined race. The strategy proposal server 60 inputs the generated strategy generation instruction into the generating AI model, generates racing strategy information corresponding to the strategy generation instruction, and transmits the generated racing strategy information to the information terminal 50. The information terminal 50 displays and outputs the received racing strategy information. As a result, the strategy proposal system 100 can propose a racing strategy to win against competitors in a race based on information about the target athlete 10's past races by utilizing the generating AI model. Therefore, the strategy proposal system 100 can smoothly and effectively support communication in discussions about racing strategies with the target athlete 10.

[0017] The information terminal 50 may be an information terminal such as a smartphone or tablet, or it may be a personal computer. The information terminal 50 may be a terminal used by the target player 10, or it may be a terminal used by the target player 10's coach, etc. The strategy proposal server 60 is a server computer connected to the internet that sends and receives data with the information terminal 50. The information terminal 50 and the strategy proposal server 60 may be composed of computers consisting of a CPU (Central Processing Unit), GPU (Graphics Processing Unit), RAM (Random Access Memory), ROM (Read Only Memory), auxiliary storage device, communication device, etc. The information terminal 50 and the strategy proposal server 60 may each be composed of separate computers, or they may be composed of a single computer or information terminal such as a smartphone or tablet that combines the functions of both. In this embodiment, an example of implementation with separate computers will be described.

[0018] The target athlete information includes data and input information (hereinafter collectively referred to as "past running information") recorded regarding the target athlete 10's past runs. The target athlete information may further include physical information such as the target athlete 10's height, weight, body fat percentage, and bone density. The strategy proposal system 100 of this embodiment includes a measuring device 20 for acquiring target athlete information, which includes a wristwatch-type device 12 that the target athlete 10 can wear while running, a waist-mounted device 14, and an information terminal-type device 16. The information terminal-type device 16 and the information terminal 50 may each be composed of separate information terminals, or they may be composed of a single information terminal that combines the functions of both.

[0019] The wristwatch-type device 12 is a sports watch or smartwatch that acquires location information, movement information, etc. The wristwatch-type device 12 includes sensors such as a positioning module, motion sensor, heart rate sensor, and barometer, and acquires information such as date and time, location coordinates, altitude, heart rate, heart rate variability, temperature, and running cadence (number of steps per unit time). The motion sensor basically consists of an inertial sensor that combines an accelerometer and a gyroscope. The waist-worn device 14 is an electronic device that is worn near the waist of the target athlete 10 to acquire location information, movement information, etc. The information terminal-type device 16 is a portable information terminal such as a smartphone that is held by the target athlete 10 in a pocket or elsewhere and acquires location information, movement information, etc.

[0020] The target athlete 10 wears one or more measuring devices 20 while running during test runs, training, races, etc. During this time, the measuring devices 20 acquire location information and movement information during or before / after running. If the target athlete 10 wears multiple measuring devices 20, they may use different devices depending on the information to be acquired, such as acquiring location information with a wristwatch-type device 12 and movement information with a waist-mounted device 14.

[0021] The measuring device 20 is not limited to devices such as a wristwatch-type device 12, a waist-worn device 14, or an information terminal-type device 16. The measuring device 20 may be a device worn on or inside the runner's shoes, or it may be a belt-type device that can be wrapped around the chest, wrist, waist, or arm of the target athlete 10 to acquire location information, movement information, and heart rate information. Furthermore, the measuring device 20 is not limited to a device worn by the target athlete 10; for example, it may be a camera that photographs the target athlete 10 while running, or it may be a device that analyzes the brain waves, blood lactate concentration, blood oxygen saturation, exhaled gas, etc., of the target athlete 10 during or before / after running.

[0022] The target athlete 10 runs while wearing at least one or all of the following measuring devices 20: for example, a wristwatch-type device 12, a waist-mounted device 14, and an information terminal-type device 16. The measuring device 20 transmits the acquired information as target athlete information to the strategy proposal server 60 via communication. However, since the communication means of the wristwatch-type device 12 and the waist-mounted device 14 of the measuring device 20 is short-range wireless communication, they do not communicate directly with the strategy proposal server 60, but rather synchronize information with the information terminal-type device 16, which also functions as an information terminal 50, and the information terminal 50 sends and receives information with the strategy proposal server 60. As a variation, the waist-mounted device 14 may first synchronize information with the wristwatch-type device 12 via short-range wireless communication, and the wristwatch-type device 12 may further synchronize information with the information terminal-type device 16 via short-range wireless communication.

[0023] The above-described explanation of each configuration of the strategic proposal system 100 is merely an example, and the strategic proposal system 100 can be implemented with various hardware and software configurations. The strategic proposal system 100 does not necessarily have to include measuring devices 20 such as a wristwatch-type device 12, a waist-worn device 14, or an information terminal-type device 16. In this case, for example, the target athlete 10 may input information about the target athlete 10, such as past running data, into an information terminal 50 such as a personal computer. Furthermore, the strategic proposal system 100 does not necessarily have to include an information terminal 50. In this case, the strategic proposal server 60 may receive the target athlete information from a device other than the information terminal 50, or it may store the target athlete information in advance. When the strategic proposal system 100 consists only of a single device, such a device is also called a strategic proposal device.

[0024] Furthermore, for example, the strategic proposal system 100 may be implemented by an application that runs on any one of the following devices: the wristwatch-type device 12, the waist-worn device 14, the information terminal 50, or the strategic proposal server 60, or by an application that runs on a combination of two or more of these devices. There may be one or more applications that run. For example, the strategic proposal system 100 may be implemented by a combination of an existing driving management application and an application prepared for this embodiment.

[0025] Figure 2 is a functional block diagram showing the various components of the strategic proposal system 100. In Figure 2, functional blocks are depicted for the measuring device 20 and the information terminal 50, each realized through the coordination of various hardware and software configurations. Therefore, it will be understood by those skilled in the art that these functional blocks can be realized in various ways using hardware alone, software alone, or a combination thereof. The measuring device 20 is composed of a combination of hardware such as a microprocessor, display device, memory, communication module, positioning module, motion sensor, and optical heart rate monitor. The information terminal 50 is composed of a combination of hardware such as a microprocessor, touch panel, memory, communication module, positioning module, and motion sensor. The functions of the measuring device 20 and the information terminal 50 will be described below.

[0026] The measuring device 20 includes a communication unit 21, a time measurement unit 22, a position measurement unit 24, a motion detection unit 26, an environmental measurement unit 27, and a calculation unit 28. For example, a waist-worn device 14 is used as the measuring device 20. The time measurement unit 22 measures the running start time, i.e., the running time from the measurement start time, by counting a timer. The position measurement unit 24 measures the current position of the target athlete 10 using position information received from a satellite positioning system by a positioning module such as a GPS (Global Positioning System) module. The motion detection unit 26 detects motion information such as the running pitch, pelvic rotation and translational movement, or impact value of the target athlete 10 using a motion sensor. The environmental measurement unit 27 acquires environmental information such as temperature, humidity, and atmospheric pressure when the target athlete 10 is running using the sports equipment 18, using a temperature sensor, humidity sensor, atmospheric pressure sensor, etc.

[0027] The calculation unit 28 may calculate new motion information and environmental information based on exercise time, location information, motion information, and environmental information. For example, the calculation unit 28 may calculate new motion information such as running time, running distance, running speed, stride, ground contact pattern, balance, and other running form information. For example, the calculation unit 28 may calculate new environmental information such as the undulations and type of the road surface during running, the slope, and the elevation. Information on the type of road surface may include, for example, paved roads, gravel roads, and mountain trails. Alternatively, the information terminal 50 may calculate the above-mentioned new motion information and environmental information instead of the calculation unit 28.

[0028] A wristwatch-type device 12 or an information terminal-type device 16 can also be used as the measuring device 20. When a wristwatch-type device 12 is used as the measuring device 20, the motion detection unit 26 of the wristwatch-type device 12 detects motion information such as the pitch of the target player 10 using a motion sensor and detects physiological indicator information such as heart rate using an optical heart rate monitor. Physiological indicator information may also be included in the motion information. When an information terminal-type device 16 is used as the measuring device 20, the motion detection unit 26 of the information terminal-type device 16 detects motion information such as the pitch of the target player 10 using a motion sensor. The information terminal 50 may also function as the information terminal-type device 16 used as the measuring device 20, in which case, for example, a single mobile device such as a smartphone may have all the functions of both the measuring device 20 and the information terminal 50.

[0029] The information terminal 50 includes an information acquisition unit 30, an input / output unit 51, and a communication unit 52. The information acquisition unit 30 receives location information, motion information, and environmental information related to the target athlete 10's running, measured or detected at each intermediate point in the run by a measuring device 20 worn by the target athlete 10, via the communication unit 52. Here, "intermediate point" refers to a point in time or distance during the run, and location information, motion information, and environmental information are measured and recorded in the measuring device 20 for each time or distance intermediate point. The information acquisition unit 30 may synchronize information with the measuring device 20 and acquire information from the measuring device 20 while the target athlete 10 is running, or it may acquire all the running information from the measuring device 20 at once after the target athlete 10 has finished running.

[0030] The input / output unit 51 receives input from the target athlete 10, etc. For example, the input / output unit 51 receives information about the shoes worn during the run through the operation of the target athlete 10, etc. The information used as assumptions for the running strategy is an example of external information, which will be described in detail later. The input / output unit 51 transmits the location information, movement information, and environmental information acquired by the measuring device 20, along with various information based on the input from the target athlete 10, etc., as target athlete information to the strategy proposal server 60 via the communication unit 52. The input / output unit 51 displays the running strategy information received from the strategy proposal server 60 on the screen. The input / output unit 51 may be composed of hardware such as a touch panel, speaker, microphone, etc. The measuring device 20 or information terminal 50 may also acquire weather information such as temperature, humidity, weather, wind direction, and wind speed corresponding to the time the target athlete 10 is running from a predetermined server, and transmit that weather information to the strategy proposal server 60 along with the environmental information.

[0031] Figure 3 is a functional block diagram showing the various functions of the strategic proposal server 60. It depicts the functional blocks of the strategic proposal server 60, which are realized through the coordination of various hardware and software configurations. Therefore, it will be understood by those skilled in the art that these functional blocks can be realized in various ways using hardware alone, software alone, or a combination thereof. The strategic proposal server 60 is composed of, for example, a combination of hardware such as a microprocessor, memory, and communication modules.

[0032] The strategy proposal server 60 comprises a communication unit 62, an acquisition unit 64, an analysis unit 70, an instruction generation unit 74, a strategy generation unit 76, and an output unit 78. The acquisition unit 64 includes a target player information acquisition unit 66 and an external information acquisition unit 68.

[0033] The target athlete information acquisition unit 66 acquires target athlete information from the information terminal 50 via the communication unit 62. As described above, the target athlete information includes the target athlete 10's past running information. The past running information included in the target athlete information acquired by the target athlete information acquisition unit 66 includes main information directly related to the target athlete 10's past running. The main information of past running information includes, for example, running speed, running time, running distance, altitude, weather, etc., during the run.

[0034] The past running information included in the target athlete information acquired by the target athlete information acquisition unit 66 further includes, in addition to the main information, at least one of the following: information on skills, information on mental aspects, and information on physical aspects. Information on skills includes, for example, running pitch, ground contact pattern, stride, balance, and other running form during running. Information on mental aspects includes, for example, perceived exercise intensity during or before / after running, and stress index during or before / after running. Perceived exercise intensity is information input based on the subjective perception of the target athlete 10. The stress index is obtained, for example, by estimation from electroencephalograms, estimation from heart rate variability, etc. Information on physical aspects includes, for example, heart rate during or before / after running, anaerobic threshold (AT) during or before / after running, and jump height measured before and after running. The anaerobic threshold can be analyzed from blood lactate concentration or exhaled gas. Of the past running information, each piece of information obtained during running may be information measured or detected at each intermediate point. In other words, the information obtained during running may include information that changes over time due to the fatigue of the subject athlete 10, etc. For example, running pitch, ground contact pattern, stride, balance, and other running form may include information that changes depending on the duration of running.

[0035] The past running information included in the target athlete information acquired by the target athlete information acquisition unit 66 may include information about the shoes worn by the target athlete 10 during the run. Different shoes have different characteristics; for example, they may be more suitable for a high-cadence running style or a long-stride running style, or they may have high or low rebound properties. Such shoe characteristics can influence the skill, mental, and physical aspects of the past running information.

[0036] The target player information acquisition unit 66 does not necessarily have to acquire the target player information from the information terminal 50. The target player information may be stored in advance within the strategy proposal server 60, or it may be acquired from a server other than the strategy proposal server 60. For example, the target player information acquisition unit 66 may acquire the target player information via the communication unit 62 from a running management server (not shown) that manages information related to the running of the target player 10. The target player information acquisition unit 66 may store the acquired target player information in association with the identification information of the target player 10. This allows the strategy proposal system 100 to propose running strategies for multiple users, including the target player 10.

[0037] The analysis unit 70 analyzes the characteristics of the target player based on the target player information acquired by the target player information acquisition unit 66. The target player characteristics are the characteristics of the target player 10 that relate to running performance. The analysis unit 70 analyzes the target player characteristics using a predetermined analysis model 72. The analysis model 72 is not particularly limited and may be, for example, a model based on statistical methods such as regression models, a deep learning model, a large language model (LLM), etc. The analysis unit 70 may extract the information necessary to analyze the target player characteristics from the target player information in advance and then analyze the target player characteristics. This ensures that the target player characteristics can be appropriately analyzed even if the target player information acquired by the target player information acquisition unit 66 contains information that is unnecessary for the analysis.

[0038] The analysis unit 70 analyzes the characteristics of the target athlete 10, which include information on the degree to which the athlete's running ability is affected by the running conditions that influence the athlete's running ability. Specifically, the analysis unit 70 analyzes the characteristics of the target athlete 10, which include information on the degree to which at least one of the running conditions, such as pace fluctuations, course elevation fluctuations, and weather fluctuations, affects at least one of the skill, mental, and physical aspects of the athlete's running ability. The skill, mental, and physical aspects of the athlete 10's running ability are the running abilities related to the information on skill, mental, and physical aspects included in the athlete information acquired by the athlete information acquisition unit 66, respectively. The degree to which the athlete 10's running ability is affected by the running conditions is a characteristic that differs from athlete 10 to athlete 10. Pace fluctuation refers to fluctuations in running pace, and information on the degree of influence of pace fluctuation indicates how much the running ability of the subject athlete 10 is affected when the subject athlete 10 runs while changing their running pace compared to when the subject athlete 10 runs at a constant pace. Information on the degree of influence of course elevation fluctuation indicates how much the running ability of the subject athlete 10 is affected when the elevation fluctuations are large compared to when the course the subject athlete 10 is running on is flat. Information on the degree of influence of weather fluctuation indicates how much the running ability of the subject athlete 10 is affected when the weather changes while the subject athlete 10 is running compared to when the weather does not change. Weather fluctuation refers to fluctuations in at least one of the following: weather, temperature, humidity, wind direction, wind speed, temperature, and humidity.

[0039] The information regarding the degree to which a player's running ability is affected by running conditions can be expressed in any way. For example, it could be information defining the correspondence between each parameter related to running conditions and each parameter related to running ability, or it could be information that evaluates the skill, mental, and physical aspects of running ability in two or more stages, indicating whether they are easily affected or not, for each running condition. Furthermore, the information regarding the degree to which a player's running ability is affected by running conditions can be expressed in text, mathematical formulas, images, or any combination thereof.

[0040] The analysis unit 70 may perform an analysis using information that serves as a predetermined set of criteria. When the analysis unit 70 performs an analysis using information that serves as a set of criteria, such criteria may be based on known scientific data published in papers, etc., or on new analyses. The criteria may also include indicators related to the main information contained in the target player information. The criteria may further include indicators related to at least one of the following: information on skills, information on mental aspects, and information on physical aspects contained in the target player information.

[0041] The external information acquisition unit 68 acquires information as external information, specifically information about a predetermined race in which the target athlete 10 is scheduled to participate (hereinafter also simply referred to as "race information"). The external information acquisition unit 68 may acquire the race information from the information terminal 50 via the communication unit 62, or from any server connected to the Internet via the communication unit 62.

[0042] Race information includes at least one of the following: race course information, weather information related to the race, race time setting information, and information on virtual competitors in the race. Race course information includes, for example, elevation changes, slopes, and road surface undulations and types for each section of the race course. Race course information may also be location information indicating the course route, such as GPX (GPS eXchange Format). Weather information related to the race is information on the weather expected on the day the race is held, and may be the latest forecast information from weather satellite communications, or information on the weather trends on the day the race is held in the region where the race is held. If the race is held repeatedly at the same location and at the same time, it may also be weather information from past races corresponding to the race in question.

[0043] Information on virtual competitors in a race may include information on participants in past races corresponding to the current race, or information on participants scheduled to participate in the current race. Information on participants in past races may include, for example, information on lap times for each section of the race for top finishers. Information on participants scheduled to participate in a race may include, for example, characteristics related to the performance of the participants. Characteristics related to the performance of the participants may be information obtained by analyzing the same characteristics as those of the target participants described above for the participants. If a participant has participated in a past race corresponding to the current race, the information on that participant may include information on their lap times for each section of the race. Note that the information on virtual competitors may differ for each of the multiple competitors participating in the race.

[0044] The instruction generation unit 74 generates strategy generation instructions based on the characteristics of the target player. Here, the characteristics of the target player used to generate the strategy generation instructions are the information analyzed by the analysis unit 70 based on the target player information, as described above. In other words, the instruction generation unit 74 also generates strategy generation instructions based on the target player information. In generating strategy generation instructions, the instruction generation unit 74 also generates strategy generation instructions based on the race information acquired by the external information acquisition unit 68.

[0045] The strategy generation instruction is an instruction for causing the driving strategy information generation AI model 92 to output. Here, before explaining the details of the instruction generation unit 74, the generation AI model 92 will be explained. The generation AI model 92 is provided in a generation AI server 90 outside the strategy proposal system 100. The generation AI model 92 is a model that generates and outputs information according to the generation instruction when a generation instruction such as a prompt is input. The generation AI model 92 is, for example, a generation model such as a large multi-modal model (LMM: Large Multimodal Model) or a large language model (LLM: Large Language Model).

[0046] Returning to the explanation of the instruction generation unit 74. The strategy generation instruction generated by the instruction generation unit 74 is an example of a generation instruction such as a prompt input to the generation AI model 92. The strategy generation instruction includes, for example, at least any one of character information, mathematical formulas, images, and any combination thereof. The driving strategy information for which the strategy generation instruction instructs the generation AI model 92 to output includes information indicating how to drive in order for the target racer 10 to win against the virtual competitor in a predetermined race. Specifically, the driving strategy information is information corresponding to the virtual competitor in a predetermined race and includes, for example, information on the recommended pace for each driving section. The recommended pace for each driving section may be a pace set so that the probability of the target racer 10 winning against the virtual competitor is high in relation to the assumed pace of the virtual competitor.

[0047] The instruction generation unit 74 may generate a strategy generation instruction so as to specify, as driving conditions that affect the driving ability of the target racer 10 based on the target racer characteristics, at least one of the race conditions and the influence on the driving ability of the target racer 10 by the assumed pace of the virtual competitor, and reflect the same in the driving strategy information. Here, the race conditions are information including, for example, information on the race course and weather information expected on the race day. The information on the race course may include information on the elevation change of the course. The weather information expected on the race day may include information on the expected weather change. The assumed pace of the virtual competitor includes, for example, information on the driving pace for each driving section expected to be driven by the virtual competitor in the race. When there are a plurality of virtual competitors, it may be the assumed pace of the leading group. Thereby, since the influence of driving conditions related to the race, such as the race conditions and the assumed pace of the virtual competitor, on the driving ability of the target racer 10 can be reflected in the driving strategy, the strategy proposal system 100 can propose a more accurate driving strategy.

[0048] The driving strategy information for which the strategy generation instruction instructs the output to the generation AI model 92 may include text information indicating at least any one of improvement, maintenance, and suppression of the driving pace for each driving section. The text information indicating the improvement of the driving pace is information recommending to improve the driving pace, and may also mean an increase, acceleration, or increase of the driving pace. The text information indicating the suppression of the driving pace is information recommending to suppress the driving pace, and may also mean a decrease, deceleration, or decrease of the driving pace. The text information for each driving section may be information related to the recommended pace set so that the probability of the target racer 10 winning against the virtual competitor becomes high in relation to the assumed pace of the virtual competitor. Thereby, since the strategy proposal system 100 can show the specific content related to the recommended pace for each driving section in text information, it can more smoothly support the communication in the discussion of the driving strategy with the target racer 10.

[0049] The strategy generation unit 76 inputs a strategy generation instruction to the generation AI model 92 via the communication unit 62 and generates driving strategy information corresponding to the strategy generation instruction by acquiring driving strategy information output from the generation AI model 92. The acquired driving strategy information includes, for example, at least one of text information, mathematical formulas, images, and any combination thereof.

[0050] The output unit 78 outputs driving strategy information to the information terminal 50 via the communication unit 62. The information terminal 50 displays the received driving strategy information on its screen. This allows the system to propose a driving strategy to help the target athlete win against their opponents in a race, based on their past driving data.

[0051] The generating AI model 92 may be a model that has already learned the aforementioned race information. For example, if the race information is publicly known, the generating AI model 92 may have already learned the race information stored on any server connected to the internet. Alternatively, the generating AI model 92 may be a model that has been fine-tuned in advance using the race information, or it may be optimized to reference the race information using RAG (Retrieval Augmented Generation) or the like. In this case, the instruction generation unit 74 may generate strategy generation instructions without using the race information acquired by the external information acquisition unit 68 when generating strategy generation instructions. Also, the strategy proposal server 60 does not necessarily have an external information acquisition unit 68. When generating strategy generation instructions without using the race information acquired by the external information acquisition unit 68, the instruction generation unit 74 may include information that identifies the race, such as the race name, in the strategy generation instructions. However, by using the race information acquired by the external information acquisition unit 68 when generating strategy generation instructions, the accuracy of the race information can be improved, and the accuracy of the driving strategy information can be improved.

[0052] Similarly, the generating AI model 92 may be a model that has already learned the information of the judgment criteria used by the analysis unit 70 described above. For example, if the information of the judgment criteria used by the analysis unit 70 is publicly known, the generating AI model 92 may have already learned the information of the judgment criteria stored on any server connected to the internet. The generating AI model 92 may also be a model that has been fine-tuned in advance using the information of the judgment criteria, or it may be optimized to allow reference to the information of the judgment criteria by RAG or the like. In this case, the instruction generation unit 74 may generate the strategy generation instruction by directly using the target player information acquired by the target player information acquisition unit 66, without using the target player characteristics analyzed by the analysis unit 70. Also, the strategy proposal server 60 does not necessarily have an analysis unit 70. When generating a strategy generation instruction without using the target player characteristics analyzed by the analysis unit 70, the instruction generation unit 74 may, for example, include information instructing the generating AI model 92 to analyze the target player characteristics in the strategy generation instruction. However, by using the target player characteristics analyzed by the analysis unit 70 when generating strategy generation instructions, the instruction generation unit 74 can improve the accuracy of the analysis results of the target player 10, thereby improving the accuracy of the running strategy information.

[0053] Figure 4 is a flowchart illustrating the process of strategy proposal processing in the strategy proposal server 60. The target player information acquisition unit 66 acquires target player information (S10). The analysis unit 70 analyzes the characteristics of the target player using the analysis model 72 based on the target player information acquired by the target player information acquisition unit 66 (S12). The external information acquisition unit 68 acquires race information (S14). The instruction generation unit 74 generates strategy generation instructions based on the target player characteristics and race information (S16). The strategy generation unit 76 inputs the strategy generation instructions to the generation AI model 92 and acquires the running strategy information output from the generation AI model 92 to generate running strategy information corresponding to the strategy generation instructions (S18). The output unit 78 outputs the running strategy information (S20).

[0054] The strategy suggestion server 60 does not have to perform the process in step S14. In this case, in step S16, the instruction generation unit 74 can be interpreted as generating a strategy generation instruction based on the target player characteristics. The strategy suggestion server 60 does not have to perform the process in step S12. In this case, in step S16, the instruction generation unit 74 can be interpreted as generating a strategy generation instruction based on the target player information and race information. The strategy suggestion server 60 does not have to perform the processes in steps S12 and S14. In this case, in step S16, the instruction generation unit 74 can be interpreted as generating a strategy generation instruction based on the target player information. Note that the order of processing shown in Figure 4 is merely an example. In particular, the order of processing in steps S12 and S14 can be either first or second.

[0055] The following will explain various types of information related to the strategy proposal system 100 with specific examples. Figure 5 shows an example of a screen that displays and outputs information about the target player. The example shown in Figure 5 is information obtained when player X, an example of the target player 10, performed a predetermined test run. The test run includes running at a constant pace and running at a variable pace. The test run also includes sections where the player runs in a straight line and sections where the player runs around corners. The first display item 102 includes a facial image of player X and text information indicating player X's name. The second display item 104 is a table showing the running pitch range when running at a constant pace and a variable pace in the straight section and the corner section, respectively. The third display item 106 is a graph showing the heart rate at both the constant pace and the variable pace. The fourth display item 108 is a graph showing the perceived exercise intensity at both the constant pace and the variable pace. The fifth display item 110 is a table showing which of the following items—"form" (an example of skill-related aspects), "mind" (an example of mental aspects), and "body" (an example of physical aspects)—is more susceptible to the influence of a constant pace or a variable pace. The sixth display item 112 is a graph showing the lactic acid concentration before and after running at a constant pace and a variable pace, respectively. The seventh display item 114 is a graph showing the jump height before and after running at a constant pace and a variable pace, respectively.

[0056] Figure 6 shows an example of a screen that displays the characteristics of the target player. The example shown in Figure 6 is the result of analysis performed by the analysis unit 70 based on the target player information for player X shown in Figure 5. The eighth display item 120 is a facial image of player X. The ninth display item 122 is text information indicating the degree to which pace fluctuations affect the physical aspect of player X's running ability. The tenth display item 124 is text information indicating the degree to which pace fluctuations affect the mental aspect of player X's running ability. The eleventh display item 126 is text information indicating the degree to which pace fluctuations affect the skill aspect of player X's running ability. The twelfth display item 128 is text information indicating advice based on the degree to which pace fluctuations affect each aspect of player X's running ability.

[0057] Figure 7 shows a first example of race course information included in race information. The example shown in Figure 7 is race course information included in the race information of a given first marathon, and includes an elevation graph 130 and an explanatory text 132. The elevation graph 130 is a graph with each point from the start to the finish of the first marathon race course on the horizontal axis and elevation on the vertical axis, and is, for example, image information. The explanatory text 132 is text information that expresses the elevation changes for each running section of the first marathon race course. In addition to the information shown in Figure 7, the race information for the first marathon further includes information on virtual competitors.

[0058] Figure 8 shows a first example of a screen that displays and outputs running strategy information. The example shown in Figure 8 is running strategy information generated by the strategy generation unit 76 using strategy generation instructions generated by the instruction generation unit 74 based on the target athlete characteristics for athlete X shown in Figure 6 and the race information of the first marathon described above. The 13th display item 140 is a facial image of athlete X. The 14th display item 142 is text information indicating information that is assumed when generating the running strategy information. The 14th display item 142 includes text information indicating the set time and text information indicating virtual competitors. In this example, the text information indicating virtual competitors indicates that half of the competitors in the first marathon are good at maintaining a constant pace, and the other half are good at maintaining a variable pace. The 15th display item 144 is text information that specifically indicates the running strategy for each running section. In the example of the 15th display item 144, it includes information on the specific recommended pace for each running section (for example, "3:00 / km"). Display item 146 includes information on the total time expected when driving according to the driving strategy, and textual information summarizing the driving strategy.

[0059] Figure 9 shows a second example of race course information included in race information. The example shown in Figure 9 is race course information included in the race information for a given second marathon, and includes elevation graph 130A. The course for the second marathon is relatively flatter than the course for the first marathon shown in Figure 7.

[0060] Figure 10 shows an example of information on virtual competitors included in race information. The example shown in Figure 10 is information on virtual competitors included in the race information for the second marathon, and is a table showing the total time and time for each running section for the top 5 runners from past second marathon races.

[0061] Figure 11 shows a second example of a screen that displays and outputs running strategy information. The example shown in Figure 11 is running strategy information generated by the strategy generation unit 76 using strategy generation instructions generated by the instruction generation unit 74 based on the target athlete characteristics for athlete X shown in Figure 6 and the race information of the second marathon described above. The 16th display item 140A is a facial image of athlete X. The 17th display item 142A is text information indicating information that is assumed when generating the running strategy information. The 17th display item 142A includes text information indicating a virtual competitor. The 18th display item 144A is text information that specifically shows the running strategy for each running section. In the example of the 18th display item 144A, it does not include information on specific recommended paces for each running section, but it does include text information indicating at least one of improving, maintaining, or suppressing the running pace for each running section. The 19th display item 146A is text information that concisely expresses the key points of the running strategy. The 20th display item 148A is an elevation difference graph of the race course for the second marathon. Display item 148A, which assigns symbols to multiple travel sections, clearly indicates the correspondence with each travel section in display item 144A.

[0062] Figure 12 shows a third example of a screen that displays and outputs running strategy information. The running strategy information in the example shown in Figure 12 includes information on multiple recommended paces corresponding to multiple simulations of the race situation of the target athlete 10. Specifically, the running strategy information shown in Figure 12 includes running strategy 150 for the first running section, running strategy 152 for the second running section, running strategies 154A and 154B for the third running section, running strategies 156A and 156B for the fourth running section, and running strategies 158A to 158D for the fifth running section. Each running strategy for each running section includes information on at least the recommended pace. Some running strategies (the second and fourth running sections) further include textual information that specifically indicates the running strategy. In the third running section, as shown in running strategies 154A and 154B, it includes information on two recommended paces corresponding to two simulations of the race situation in the second running section. Similarly, in the fifth running section, as shown in running strategies 158A to 158D, information on a total of four recommended paces is included, corresponding to two simulations for each of the two race conditions in the fourth running section.

[0063] As shown in Figure 12, in order to generate running strategy information, the instruction generation unit 74 may generate a strategy generation instruction that causes the AI ​​model 92 to output running strategy information that includes information on multiple recommended paces corresponding to multiple simulations of the target athlete 10's race situation for at least one running section. As a result, the strategy proposal system 100 can propose running strategies corresponding to multiple simulations, and can propose running strategies that take into account the tactics against competitors.

[0064] [Second Embodiment] Figure 13 is a functional block diagram showing the functions of the instruction generation server 60A included in the instruction generation system 200 according to the second embodiment. The instruction generation system 200 comprises an information terminal 50 and an instruction generation server 60A. The instruction generation system 200 optionally includes a wristwatch-type device 12, a waist-worn device 14, and an information terminal-type device 16 as measuring devices 20 for acquiring target player information. The wristwatch-type device 12, waist-worn device 14, information terminal-type device 16, measuring devices 20, and information terminal 50 are the same as their respective configurations in the first embodiment, so their descriptions are omitted.

[0065] The instruction generation server 60A comprises a communication unit 62, an acquisition unit 64, an analysis unit 70, an instruction generation unit 74, and an output unit 78. The acquisition unit 64 includes a target player information acquisition unit 66 and an external information acquisition unit 68. In other words, the instruction generation server 60A does not necessarily need to include the strategy generation unit 76 in the strategy proposal server 60 of the first embodiment, but the other configurations of the instruction generation server 60A are the same as those of the strategy proposal server 60.

[0066] The output unit 78 of the instruction generation server 60A outputs the strategy generation instructions generated by the instruction generation unit 74 to the information terminal 50 via the communication unit 62. As a result, the instruction generation system 200 can generate instructions to propose a racing strategy to win against competitors in a race, based on the target athlete's past racing information, using the generation AI model 92. Therefore, the instruction generation system 200 can smoothly and effectively support communication in discussions about racing strategies with the target athlete 10 by utilizing the generation AI model 92.

[0067] The present disclosure has been described above based on embodiments. The embodiments are illustrative, and it will be understood by those skilled in the art that various modifications are possible in the combination of their components and processing processes, and that such modifications are also within the scope of the present disclosure. Furthermore, the above-described embodiments can be generalized to obtain the following embodiments.

[0068] [Aspect 1] A strategy proposal system that proposes running strategies in a running race using a generating AI model, wherein the generating AI model is a model that generates and outputs information corresponding to a generation instruction when a generation instruction is input, and comprises: an acquisition unit that acquires target athlete information including information on the target athlete's past running; an instruction generation unit that generates a strategy generation instruction to cause the generating AI model to output running strategy information corresponding to a virtual competitor in a predetermined race based on the target athlete information; a strategy generation unit that generates running strategy information corresponding to the strategy generation instruction by inputting the strategy generation instruction to the generating AI model; and an output unit that outputs the running strategy information.

[0069] According to the strategy proposal system of Embodiment 1, by utilizing a generative AI model, it is possible to propose a running strategy to win against competitors in a running race based on information about the target athlete's past running. Therefore, according to the strategy proposal system of Embodiment 1, communication in discussions about running strategies with the target athlete can be smoothly and effectively supported.

[0070] [Aspect 2] The strategy proposal system according to aspect 1, further comprising an analysis unit that analyzes the characteristics of the target player, including information on the degree to which the running ability is affected by running conditions that affect the running ability of the target player, using a predetermined analysis model based on the target player information, and the instruction generation unit may generate the strategy generation instruction so as to specify the effect on the running ability of the target player by at least one of the conditions of the race and the assumed pace of the virtual competitor as running conditions, based on the target player characteristics, and reflect this in the running strategy information.

[0071] According to the strategy proposal system of Embodiment 2, the influence of race-related running conditions on the target athlete's running ability can be reflected in the running strategy, thus enabling the proposal of a more accurate running strategy.

[0072] [Aspect 3] The strategy proposal system according to aspect 2, wherein the analysis unit may, in analyzing the characteristics of the target player, analyze the characteristics of the target player including information on the degree to which at least one of the running conditions, such as pace fluctuations, course elevation fluctuations, and weather fluctuations, has an effect on at least one of the running abilities of the target player, such as skill, mental, and physical aspects.

[0073] According to the strategy proposal system of embodiment 3, it is possible to analyze the degree to which predetermined factors of running conditions have an influence on any aspect of the running ability of the target athlete, thereby enabling the proposal of a more accurate running strategy.

[0074] [Aspect 4] The strategy proposal system according to any one of aspects 1 to 3, wherein the instruction generation unit may generate a strategy generation instruction to cause the generating AI model to output driving strategy information, which includes text information indicating at least one of improving, maintaining, or suppressing the driving pace for each driving section, as the driving strategy information.

[0075] According to the strategy proposal system of embodiment 4, specific details related to the recommended pace for each running section can be shown in text information, thus facilitating smoother communication in discussions about running strategies with the target athlete.

[0076] [Aspect 5] The strategy proposal system according to any one of aspects 1 to 4, wherein the instruction generation unit may generate a strategy generation instruction that causes the generating AI model to output running strategy information including information on multiple recommended paces corresponding to multiple simulations of the target athlete's race situation in at least one running section.

[0077] According to the strategy proposal system of embodiment 5, it is possible to propose driving strategies corresponding to multiple simulations, and therefore it is possible to propose driving strategies that take into account the maneuvering with competitors.

[0078] [Aspect 6] An instruction generation system for generating instructions to an AI model for proposing a running strategy in a running race, wherein the AI ​​model is a model that generates and outputs information corresponding to a generation instruction when a generation instruction is input, and comprises: an acquisition unit that acquires target athlete information including information on the target athlete's past running; and an instruction generation unit that generates a strategy generation instruction to cause the AI ​​model to output running strategy information corresponding to a virtual competitor in a predetermined race, based on the target athlete information.

[0079] According to the instruction generation system of embodiment 6, instructions can be generated using a generative AI model to propose a running strategy for winning against competitors in a running race, based on information about the target athlete's past running. Therefore, according to the instruction generation system of embodiment 6, by utilizing the generative AI model, communication in discussions about running strategies with the target athlete can be smoothly and effectively supported.

[0080] [Aspect 7] A strategy proposal method for proposing a running strategy in a running race using a generating AI model, wherein the generating AI model is a model that generates and outputs information corresponding to a generation instruction when a generation instruction is input, and the method includes: a process in which a computer acquires target athlete information including information on the target athlete's past running; a process in which the computer generates a strategy generation instruction based on the target athlete information to cause the generating AI model to output running strategy information corresponding to a virtual competitor in a predetermined race; a process in which the computer generates running strategy information corresponding to the strategy generation instruction by inputting the strategy generation instruction into the generating AI model; and a process in which the computer outputs the running strategy information.

[0081] According to the strategy proposal method of Embodiment 7, by utilizing a generative AI model, it is possible to propose a running strategy for winning against competitors in a running race based on information about the target athlete's past running. Therefore, according to the strategy proposal method of Embodiment 7, communication in discussions about running strategies with the target athlete can be smoothly and effectively supported.

[0082] [Aspect 8] A strategy proposal program for a computer that proposes a running strategy in a running race using a generation AI model, wherein the generation AI model is a model that generates and outputs information corresponding to a generation instruction when a generation instruction is input, and has the following functions: a function to acquire target athlete information including information on the target athlete's past running; a function to generate a strategy generation instruction based on the target athlete information to cause the generation AI model to output running strategy information corresponding to a virtual competitor in a predetermined race; a function to input the strategy generation instruction to the generation AI model to generate running strategy information corresponding to the strategy generation instruction; and a function to output the running strategy information.

[0083] According to the strategy proposal program of Embodiment 8, by utilizing a generative AI model, it is possible to propose a running strategy to win against competitors in a running race based on information about the target athlete's past running. Therefore, according to the strategy proposal program of Embodiment 8, communication in discussions about running strategies with the target athlete can be smoothly and effectively supported.

[0084] This disclosure relates to a technology that proposes running strategies in running races.

[0085] 10 Target players, 60 Strategy proposal server, 64 Acquisition unit, 66 Target player information acquisition unit, 68 External information acquisition unit, 70 Analysis unit, 72 Analysis model, 74 Instruction generation unit, 76 Strategy generation unit, 78 Output unit, 90 Generation AI server, 92 Generation AI model, 100 Strategy proposal system, 200 Instruction generation system.

Claims

1. A strategy proposal system that proposes running strategies in a running race using a generating AI model, wherein the generating AI model is a model that generates and outputs information corresponding to a generation instruction when a generation instruction is input, and comprises: an acquisition unit that acquires target athlete information including information on the target athlete's past running; an instruction generation unit that generates a strategy generation instruction to cause the generating AI model to output running strategy information corresponding to a virtual competitor in a predetermined race based on the target athlete information; a strategy generation unit that generates running strategy information corresponding to the strategy generation instruction by inputting the strategy generation instruction to the generating AI model; and an output unit that outputs the running strategy information.

2. The strategy proposal system according to claim 1, further comprising an analysis unit that analyzes, based on the target athlete information, the characteristics of the target athlete, including information on the degree to which the running ability is affected by running conditions that affect the running ability of the target athlete, using a predetermined analysis model, wherein the instruction generation unit generates a strategy generation instruction based on the target athlete characteristics, specifying as the running conditions the influence of at least one of the race conditions and the assumed pace of the virtual competitor on the target athlete's running ability, and reflecting this in the running strategy information.

3. The strategy proposal system according to claim 2, wherein the analysis unit analyzes the characteristics of the target player, including information on the degree to which at least one of the running conditions, such as pace fluctuations, course elevation fluctuations, and weather fluctuations, has an effect on at least one of the running abilities of the target player, such as skill, mental, and physical aspects.

4. The strategy proposal system according to claim 1, wherein the instruction generation unit generates a strategy generation instruction for causing the generating AI model to output driving strategy information, which includes text information indicating at least one of improving, maintaining, or suppressing the driving pace for each driving section, as the driving strategy information.

5. The strategy proposal system according to any one of claims 1 to 4, wherein the instruction generation unit generates a strategy generation instruction for outputting running strategy information to the generating AI model, which includes information on multiple recommended paces corresponding to multiple simulations of the target athlete's race situation in at least one running section.

6. An instruction generation system for generating instructions to an AI model for proposing running strategies in a running race, wherein the AI ​​model is a model that generates and outputs information corresponding to a generation instruction when a generation instruction is input, and comprises: an acquisition unit that acquires target athlete information including information on the target athlete's past running; and an instruction generation unit that generates strategy generation instructions to cause the AI ​​model to output running strategy information corresponding to a virtual competitor in a predetermined race, based on the target athlete information.

7. A strategy proposal method for proposing running strategies in a running race using a generative AI model, wherein the generative AI model is a model that generates and outputs information corresponding to a generation instruction when a generation instruction is input, and the method includes: a process in which a computer acquires target athlete information including information on the target athlete's past running; a process in which the computer generates a strategy generation instruction based on the target athlete information to cause the generative AI model to output running strategy information corresponding to a virtual competitor in a predetermined race; a process in which the computer generates running strategy information corresponding to the strategy generation instruction by inputting the strategy generation instruction into the generative AI model; and a process in which the computer outputs the running strategy information.

8. A strategy proposal program for a computer that proposes running strategies in a running race using a generation AI model, wherein the generation AI model is a model that generates and outputs information corresponding to a generation instruction when a generation instruction is input, and has the following functions: a function to acquire target athlete information including information on the target athlete's past running; a function to generate strategy generation instructions based on the target athlete information to cause the generation AI model to output running strategy information corresponding to a virtual competitor in a predetermined race; a function to generate running strategy information corresponding to the strategy generation instruction by inputting the strategy generation instruction into the generation AI model; and a function to output the running strategy information.

Images