Apparatus and method for predicting robot delivery time and robot delivery route based on crowd estimation

Abstract

This invention provides a delivery time and delivery route prediction device and method based on crowd estimation that determines the route with the shortest delivery time as the delivery route. [Solution] A method for predicting delivery time and delivery route based on crowd estimation includes the steps of: setting up multiple routes that can be traveled to a destination; collecting crowd estimation information for the multiple routes through sensors on the multiple routes; calculating traffic and latency for the multiple routes based on the crowd estimation information; predicting delivery times for the multiple routes to the destination based on the traffic and latency; and determining the route with the shortest delivery time among the multiple routes as the delivery route.

Description

Technical Field

[0001] The present invention relates to a delivery time and delivery route prediction apparatus and method based on crowd estimation, and more particularly, to a delivery time and delivery route prediction apparatus and method based on crowd estimation in an indoor robot delivery service.

Background Art

[0002] In the case of outdoor delivery, variables are different from the prediction of indoor unmanned delivery time such as road conditions and driver behavior. In the case of indoor robot delivery, although it is possible to grasp the change in elevator stop / load to predict the inter-floor movement time, there is a problem that the possibility of robot boarding according to the congestion level is not considered.

[0003] Particularly, when the delivery robot uses the same elevator or entrance / exit door as people, due to traffic, it may be impossible to board the elevator or pass through the entrance / exit door at the required time.

Summary of the Invention

Problems to be Solved by the Invention

[0004] An object of an embodiment of the present invention is to count the number of people or objects on the robot movement path using an elevator, entrance / exit gate, CCTV, etc. for an indoor robot delivery service, grasp the density of a specific section, calculate traffic and latency, predict the delivery time based on the calculated traffic and latency, and determine the route with the shortest delivery time as the delivery route, and provide a delivery time and delivery route prediction apparatus and method based on crowd estimation.

Means for Solving the Problems

[0005] In one embodiment, the method for predicting delivery time and delivery route based on crowd estimation includes the steps of: setting up a plurality of routes that can be traveled to a destination; collecting crowd estimation information for the plurality of routes through sensors on the plurality of routes; calculating traffic and latency for the plurality of routes based on the crowd estimation information; predicting the delivery time for the plurality of routes to the destination based on the traffic and latency; and determining the route with the shortest delivery time among the plurality of routes as the delivery route.

[0006] The step of setting up multiple routes that can be taken to the destination may include, upon receiving a delivery request, receiving a saved travel route plan from the robot and determining multiple routes to the destination based on the travel route plan, such that the difference in travel time between the routes does not exceed a critical level.

[0007] The aforementioned sensors may include camera sensors, including CCTV sensors, thermal imaging sensors, elevator sensors, robot sensors, and entrance / exit gate sensors.

[0008] The step of collecting crowd estimation information for the plurality of routes through sensors along the plurality of routes may include the step of collecting people counting information for the plurality of routes through the camera sensor and the thermal image sensor, respectively.

[0009] The step of collecting crowd estimation information for the plurality of routes through sensors along the plurality of routes may include the step of collecting the density of a specific section, including the elevator section and the entrance / exit gate section, on the route based on entrance / exit log data obtained from the elevator sensor and the entrance / exit gate sensor.

[0010] The step of calculating traffic and latency for each of the multiple routes based on the crowd estimation information may include the step of calculating the traffic based on the people counting information and the density of the specific section and storing it in a database.

[0011] The step of calculating traffic and latency for each of the multiple routes based on the crowd estimation information may include a step of calculating the latency based on the time between the time when the crowd estimation information is collected and the time when the traffic is calculated using the crowd estimation information.

[0012] The step of predicting the delivery time for each of the multiple routes to the destination based on the traffic and latency may include the step of predicting the delivery time by combining the required time calculated based on the traffic and the delay time calculated based on the latency for each of the multiple routes.

[0013] After delivery is completed, the process may further include comparing the predicted delivery time with the actual delivery time for the final determined delivery route, updating the delivery time prediction model based on the comparison result if the comparison result satisfies certain criteria, and saving the comparison result to a database if the comparison result does not satisfy the certain criteria.

[0014] If the calculation of the traffic and latency is impossible due to a malfunction or error of the sensor, the step of predicting the delivery time for each of the multiple routes to the destination based on the traffic and latency further includes the step of predicting the average delivery time based on recent traffic information, including density and people counting information for a recent specific period stored in the database, and recent average latency information. The step of determining the route with the shortest delivery time among the multiple routes as the delivery route may include the steps of generating multiple delivery routes based on the average delivery time and setting priorities, and initially determining the delivery route with the first route having the highest priority, and finally determining the delivery route with the second route having the next highest priority based on traffic and latency information updated as delivery progresses.

[0015] In one embodiment, the crowd estimation-based delivery time and delivery route prediction device may include: a data analysis unit that collects crowd estimation information for multiple routes to a destination via sensors; a route analysis unit that calculates traffic and latency for multiple routes to a destination based on the crowd estimation information; a delivery time prediction unit that predicts the delivery time for each of the multiple routes to the destination based on the traffic and latency; and a delivery route determination unit that determines the route with the shortest delivery time among the multiple routes as the delivery route.

[0016] When a delivery request is received, the data analysis unit receives the travel route plan stored from the robot and, based on the travel route plan, can identify multiple routes to the destination where the difference in travel time between routes does not exceed a critical level.

[0017] The sensors may include camera sensors (including CCTV), lidar sensors, thermal imaging sensors, elevator sensors, robot sensors, and entrance / exit gate sensors.

[0018] The data analysis unit can collect people counting information along the multiple paths through the camera sensor and the thermal image sensor, respectively.

[0019] The data analysis unit can collect the density of a specific section, including the elevator section and the entrance / exit gate section, on the route based on the entrance / exit log data acquired from the elevator sensor and the entrance / exit gate sensor.

[0020] The route analysis unit can calculate the traffic based on the people counting information and the density of the specific section, and store it in a database.

[0021] The route analysis unit can calculate the latency based on the time between the time when the crowd estimation information is collected and the time when the traffic is calculated using the crowd estimation information.

[0022] The delivery time prediction unit can predict the delivery time by combining the required time calculated based on the traffic for each of the plurality of routes and the delay time calculated based on the latency.

[0023] After the delivery is completed, if the comparison result satisfies a specific criterion by comparing the predicted delivery time and the actual delivery time for the finally determined delivery route, the delivery time prediction model is updated based on the comparison result, and if the comparison result does not satisfy the specific criterion, it may further include a learning unit that stores the comparison result in a database.

[0024] When it is impossible to calculate the traffic and the latency due to a defect or error of the sensor, the delivery time prediction unit predicts an average delivery time based on the recent traffic information including the density and people counting information of a specific period stored in the database and the recent average latency information, and the delivery route determination unit generates a plurality of delivery routes based on the predicted average delivery time, sets priorities, determines the first delivery route as the route with the highest priority, and finally determines the delivery route as the second route with the next highest priority based on the traffic and latency information updated during the delivery process.

Advantages of the Invention

[0025] The delivery time and delivery route prediction device and method based on crowd estimation according to an embodiment of the present invention count the number of people or objects on the robot movement path using an elevator, an entrance / exit gate, a CCTV, etc. for an indoor robot delivery service, grasp the density of a specific section to calculate traffic and latency, predict the delivery time based on the calculated traffic and latency, and can determine the route with the shortest delivery time as the delivery route.

Brief Description of the Drawings

[0026] [Figure 1] Schematically shows a delivery time and delivery route prediction system based on crowd estimation according to an embodiment of the present invention. [Figure 2] It is a block diagram of a delivery time and delivery route prediction device based on crowd estimation according to an embodiment of the present invention. [Figure 3] It is a block diagram of a delivery time and delivery route prediction device based on crowd estimation according to an embodiment of the present invention. [Figure 4] It is a flowchart of a delivery time and delivery route prediction method based on crowd estimation according to an embodiment of the present invention. [Figure 5] It is a flowchart of a delivery time and delivery route prediction method based on crowd estimation according to an embodiment of the present invention. [Figure 6] It is a flowchart of a delivery time and delivery route prediction method based on crowd estimation according to an embodiment of the present invention. [Figure 7] It is a drawing for explaining a delivery time and delivery route prediction method based on crowd estimation according to an embodiment of the present invention. [Figure 8] It is a drawing for explaining a computing device according to an embodiment of the present invention.

Embodiments for Carrying Out the Invention

[0027] Hereinafter, referring to the attached drawings, embodiments of the present invention will be described in detail so that those having ordinary knowledge in the technical field to which the present invention belongs can easily implement it. However, the present invention can be realized in various different forms and is not limited to the embodiments described here. And in the drawings, in order to clearly explain the present invention, unnecessary parts for explanation are omitted, and similar parts throughout the specification are given similar drawing reference numerals.

[0028] In the specification and claims as a whole, when a part "includes" a component, this means, unless otherwise stated, that it may include other components rather than excluding them. Ordinal terms such as "first," "second," etc., can be used to describe a variety of components, but such components are not limited by such terms. Such terms are used solely for the purpose of distinguishing one component from another.

[0029] The terms "...part," "...device," and "...module" as used in this specification mean a unit capable of performing at least one of the functions or operations described herein, which can be implemented in hardware or circuitry, in software, or in combination of hardware or circuitry and software.

[0030] Embodiments of the present invention will be described below with reference to the drawings.

[0031] Figure 1 schematically shows a delivery time and delivery route prediction system based on crowd estimation according to one embodiment of the present invention.

[0032] Referring to Figure 1, the crowd estimation-based delivery time and delivery route prediction system includes a crowd estimation-based delivery time and delivery route prediction device 100, a delivery robot 10, an elevator 20, a CCTV 30, and an entrance / exit gate 40.

[0033] The crowd estimation-based delivery time and route prediction device 100, delivery robot 10, elevator 20, CCTV 30, and entrance / exit gate 40 can be connected via a wired / wireless network.

[0034] The crowd estimation-based delivery time and delivery route prediction device 100 predicts the delivery time of the delivery robot 10 based on crowd estimation information including people counting information and density information, and generates an optimal delivery route based on the predicted robot delivery time.

[0035] The delivery robot 10 can be embodied as a robot that performs unmanned delivery within a specific space, including inside a building.

[0036] The delivery robot 10 can move around inside the building using a saved travel path plan.

[0037] The delivery robot 10 can include various types of robotic sensors, including cameras and LiDAR.

[0038] Multiple elevators 20 can be located along the delivery routes within the building. Elevators 20 can include a variety of elevator sensors, including camera sensors, thermal imaging sensors, door open / close sensors, and weight sensors.

[0039] Multiple CCTV30 units can be present along the delivery routes within a building. CCTV30 units may include camera sensors and thermal imaging sensors.

[0040] Multiple entry / exit gates 40 can be located along the delivery routes within the building. Each entry / exit gate 40 can include various sensors to sense people or goods passing through and count the numbers.

[0041] Figures 2 and 3 are block diagrams of a delivery time and delivery route prediction device based on crowd estimation according to one embodiment of the present invention.

[0042] Referring to Figures 2 and 3, the crowd estimation-based delivery time and delivery route prediction device 100 may include a prediction unit 110, a sensor interface 120, a data analysis unit 130, a robot service interface 140, and a learning unit 150.

[0043] The crowd estimation-based delivery time and delivery route prediction device 100 can be connected via a sensor interface 120 to a variety of sensors 50, including camera sensors, thermal image sensors, lidar sensors, elevator sensors, robot sensors, and entrance / exit gate sensors, which are installed on the delivery robot 10, elevator 20, CCTV 30, and entrance / exit gate 40 shown in Figure 1.

[0044] The prediction unit 110 receives crowd estimation information from the sensor interface 120 and the data analysis unit 130, and can predict delivery time based on the traffic and latency calculated based on the received crowd estimation information.

[0045] The prediction unit 110 can determine the optimal delivery route from among multiple robot delivery routes based on the predicted delivery time.

[0046] The prediction unit 110 can estimate delivery time and delivery route using an artificial intelligence prediction model that is learned through repeated delivery time prediction and delivery route determination.

[0047] In one embodiment, when a delivery request is received, the data analysis unit 130 receives a stored travel route plan from the delivery robot.

[0048] The data analysis unit 130 can identify multiple routes to the destination where the difference in travel time between routes does not exceed a critical level, based on the travel route plan.

[0049] The data analysis unit 130 collects crowd estimation information for multiple routes to the destination via sensors. The crowd estimation information may include the degree of congestion in a specific environment calculated through cameras and sensors. The crowd estimation information may be calculated based on the number of people and density in a specific environment.

[0050] The data analysis unit 130 may include a people counting unit 131, a density analysis unit 132, and a traffic analysis unit 133.

[0051] The people counting unit 131 can collect people counting information along multiple paths via camera sensors, lidar sensors, and thermal imaging sensors. The people counting information may include the number of people identified within a specific space via camera sensors, etc.

[0052] The density analysis unit 132 can calculate the density of each of the multiple routes based on people counting information. The more people there are on a route, the higher the density.

[0053] The density analysis unit 132 can collect density data for specific sections, including elevator sections and entrance / exit gate sections, on a route based on entrance / exit log data acquired from elevator sensors and entrance / exit gate sensors.

[0054] Multiple routes may each contain multiple specific sections. The density analysis unit 132 can calculate the density of a section based on real-time entry and exit records of elevators and / or entrance / exit gates. In other words, the more entry and exit records there are, the higher the density.

[0055] For example, the density analysis unit 132 can determine that the density of a given section is higher than that of other sections if the number of entries and exits increases within a specified period of time.

[0056] For example, the density analysis unit 132 can determine that a large object or a large number of people entered or exited the section or area at the same time if a speed gate for disabled persons among the entrance / exit gates is opened, or if any one of the entrance / exit gates is open for an extended period of time.

[0057] The traffic analysis unit 133 can calculate traffic based on people counting information and the density of traffic in a specific section, and store it in a database DB.

[0058] Traffic can be proportional to the number of people and density along a route. Traffic is calculated for each route, and the estimated delivery time for routes with low traffic is shorter than the estimated delivery time for routes with high truck pick-up.

[0059] Traffic can include people counting information and rating information calculated based on density. Rating information can show the relative size of traffic between routes.

[0060] The traffic analysis unit 133 can determine, based on the number of people and density, whether it is possible to pass through elevators or entrance doors on the travel route at the time of delivery.

[0061] In other words, the traffic analysis unit 133 can detect situations of excessive traffic where the robot cannot use the elevator or entrance / exit gate due to specific events or crowd gatherings along the route, based on crowd estimation information.

[0062] The prediction unit 110 may include a route analysis unit 111, a delivery time prediction unit 112, and a delivery route determination unit 113.

[0063] The route analysis unit 111 can calculate traffic and latency for multiple routes to the destination based on crowd estimation information.

[0064] The route analysis unit 111 can calculate latency based on the time between the point in time when crowd estimation information is collected and the point in time when traffic is calculated using the crowd estimation information.

[0065] Latency can be a factor that delays the predicted delivery time for robotic delivery.

[0066] The delivery time prediction unit 112 can predict the delivery time for multiple routes to the destination based on traffic and latency.

[0067] The delivery time prediction unit 112 can predict the delivery time by combining the required time calculated based on traffic for each of the multiple routes and the delay time calculated based on latency.

[0068] If the calculation of traffic and latency is impossible due to a sensor malfunction or error, the delivery time prediction unit 112 can predict the average delivery time for the route based on recent traffic information, including density and people counting information for a specific period recently stored in the database DB, and recent average latency information.

[0069] The delivery route determination unit 113 can determine the delivery route from among multiple routes that has the shortest predicted delivery time.

[0070] If the traffic and latency calculations are impossible due to a sensor malfunction or error, the delivery route determination unit 113 can generate multiple delivery routes based on the predicted average delivery time and set priorities among the delivery routes.

[0071] The delivery route determination unit 113 determines the first route, which has the highest priority, as the initial delivery route, and can then finalize the delivery route using the second route, which has the next highest priority, based on traffic and latency information that is updated in real time as the robot delivery progresses.

[0072] In other words, the delivery route determination unit 113 first excludes routes with a large average delivery time from the delivery route. The delivery route determination unit 113 initially determines a first route, and if the delivery time of the first route increases by more than a critical threshold due to real-time traffic and latency before or during robot delivery, it can change the delivery route to a second route.

[0073] The prediction unit 110 can provide the user with information regarding the expected delivery time and delivery route via a user application or web page on the user terminal (USER).

[0074] The robot service interface 140 can transmit and request data for robot delivery services from the robot control server. For example, the robot service interface 140 can request details of item orders, cooking and preparation times.

[0075] The robot service interface 140 can provide the prediction unit 110 with various information related to orders and deliveries.

[0076] The robot service interface 140 can provide an initial estimated delivery time based on delivery information such as order breakdown and preparation time.

[0077] The robot service interface 140 can provide the learning unit 150 with various information for learning the predictive model.

[0078] After delivery is completed, the learning unit 150 compares the predicted delivery time for the final determined delivery route with the actual delivery time. If the comparison result satisfies a specific criterion, it updates the delivery time prediction model based on the comparison result. If the comparison result does not satisfy the specific criterion, it can save the comparison result to the database DB.

[0079] The database stores data analyzed for density and traffic. The database also stores data for predicted delivery times and determined delivery routes based on order information.

[0080] The database DB provides the stored data to the learning unit 150, which uses the provided data to update the predictive model in real time.

[0081] The database (DB) does not have to be located on the data analysis unit 130; it can be located separately.

[0082] Figures 4 to 6 are flowcharts of a method for predicting delivery time and delivery route based on crowd estimation according to one embodiment of the present invention.

[0083] The crowd estimation-based delivery time and delivery route prediction methods shown in Figures 4 to 6 can be performed using the crowd estimation-based delivery time and delivery route prediction device 100 shown in Figure 1.

[0084] Figure 4 is a flowchart showing the process of training a predictive model according to one embodiment.

[0085] In Figure 4, the delivery time and delivery route prediction device 100, based on crowd estimation, can collect sensor and infrastructure data at the start of delivery (step S410).

[0086] The crowd estimation-based delivery time and delivery route prediction device 100 can collect order information, sensor data on the robot's movement path, and infrastructure data through the sensor interface 120 and the robot service interface 140.

[0087] The crowd estimation-based delivery time and route prediction device 100 can analyze the density and traffic along the travel route (step S420). People counting information, density, and traffic can be stored in a database in real time.

[0088] The crowd estimation-based delivery time and delivery route prediction device 100 can collect delivery time information and delivery route information at the end of delivery along the delivery route extracted based on traffic (step S430).

[0089] The crowd estimation-based delivery time and delivery route prediction device 100 can analyze delivery times for each delivery route (step S440).

[0090] The crowd estimation-based delivery time and delivery route prediction device 100 can learn and update its prediction model based on the analysis data (step S450).

[0091] The crowd estimation-based delivery time and delivery route prediction device 100 can review the validity of the analysis data, and if it determines that the data is valid, it can update the prediction model based on that data.

[0092] The crowd estimation-based delivery time and delivery route prediction device 100 can determine the validity of the data by comparing the predicted delivery time with the actual delivery time and determining whether the difference is below a certain level.

[0093] Alternatively, the crowd estimation-based delivery time and delivery route prediction device 100 may determine that it is ineffective if the predicted delivery route and / or delivery time have abnormal values ​​that exceed general levels (predetermined).

[0094] The delivery time and delivery route prediction device 100, based on crowd estimation, saves data to a database if the data is invalid (step S460). The data saved in the database can later be used for training the prediction model or for informational purposes.

[0095] In Figure 5, the delivery time and delivery route prediction device 100, based on crowd estimation, can set up multiple possible robot movement routes to the destination (step S510).

[0096] The crowd estimation-based delivery time and delivery route prediction device 100, upon receiving a delivery request, receives a stored travel route plan from the robot and, based on the travel route plan, can identify multiple routes to the destination where the difference in travel time between routes does not exceed a critical level.

[0097] The crowd estimation-based delivery time and delivery route prediction device 100 calculates the number of people and density in specific sections, including elevators and entrance / exit gates, and can collect crowd estimation information for each robot movement path (step S520).

[0098] The crowd estimation-based delivery time and delivery route prediction device 100 can collect people counting information along multiple routes through a camera sensor and a thermal image sensor.

[0099] The crowd estimation-based delivery time and delivery route prediction device 100 can collect density data for specific sections, including elevator sections and entrance / exit gate sections, along the route, based on entrance / exit log data acquired from elevator sensors and entrance / exit gate sensors.

[0100] The crowd estimation-based delivery time and delivery route prediction device 100 can calculate traffic and latency for multiple routes based on crowd estimation information (step S530).

[0101] The crowd estimation-based delivery time and delivery route prediction device 100 can calculate traffic based on people counting information and the density of a specific section, and store it in a database.

[0102] The crowd estimation-based delivery time and delivery route prediction device 100 can calculate latency based on the time between the time when crowd estimation information is collected and the time when traffic is calculated using the crowd estimation information.

[0103] The crowd estimation-based delivery time and delivery route prediction device 100 can predict the delivery time for multiple routes to the destination based on traffic and latency (step S540).

[0104] The crowd estimation-based delivery time and delivery route prediction device 100 can predict delivery time by combining the required time calculated based on traffic for each of multiple routes and the delay time calculated based on latency.

[0105] The crowd estimation-based delivery time and delivery route prediction device 100 can determine the route with the shortest delivery time among multiple routes as the final delivery route (step S550).

[0106] The crowd estimation-based delivery time and delivery route prediction device 100 can, after delivery is completed, compare the predicted delivery time for the final determined delivery route with the actual delivery time. If the comparison result satisfies specific criteria, it can update the delivery time prediction model based on the comparison result.

[0107] The crowd estimation-based delivery time and delivery route prediction device 100 can save the comparison results to a database if the comparison results do not satisfy the specified criteria.

[0108] Figure 6 is a flowchart showing a method for predicting delivery time and delivery route based on crowd estimation according to one embodiment, in cases where traffic and latency calculation is impossible due to a sensor malfunction or error.

[0109] In Figure 6, the delivery time and delivery route prediction device 100, based on crowd estimation, can set up multiple possible robot movement routes to the destination (step S610).

[0110] The crowd estimation-based delivery time and delivery route prediction device 100 can collect recent traffic information, including density and people counting information for a specific period of time, and recent average latency information for multiple robot movement routes from a database (step S620).

[0111] The crowd estimation-based delivery time and delivery route prediction device 100 can predict delivery times based on recent traffic information and recent average latency information, generate multiple delivery routes, and set priorities (step S630).

[0112] The crowd estimation-based delivery time and route prediction device 100 first determines the delivery route with the highest priority (first route), and then can finalize the delivery route with the next priority (second route) based on traffic and latency information updated during delivery (stage S640).

[0113] Figure 7 is a diagram illustrating a method for predicting delivery time and delivery route based on crowd estimation according to one embodiment of the present invention.

[0114] In Figure 7, there are two possible delivery routes from origin A to destination B: the first delivery route PATH1 and the second delivery route PATH2.

[0115] The estimated delivery time (ETA) via the first delivery route, PATH1, is 12 minutes, taking into account both traffic (10 minutes) and latency (2 minutes). In contrast, the estimated delivery time via the second delivery route, PATH2, is 14 minutes.

[0116] The crowd estimation-based delivery time and delivery route prediction device 100 can ultimately determine a first delivery route PATH1 with a shorter predicted delivery time based on traffic and latency determined based on density. Figure 8 is a diagram illustrating a computing device according to one embodiment of the present invention.

[0117] Referring to Figure 8, the robot delivery time and robot delivery route prediction device and method based on crowd estimation according to the embodiment can be implemented using the computing device 900.

[0118] The computing device 900 may include at least one of a processor 910, memory 930, user interface input device 940, user interface output device 950, and storage device 960 that communicate via bus 920. The computing device 900 may also include a network interface 970 that is electrically connected to network 90. ​​The network interface 970 can transmit or receive signals with other devices via network 90.

[0119] The processor 910 can be implemented in various forms such as an MCU (Micro Controller Unit), AP (Application Processor), CPU (Central Processing Unit), GPU (Graphics Processing Unit), or NPU (Neural Processing Unit), and can be any semiconductor device that executes instructions stored in the memory 930 or storage device 960. The processor 910 can be configured to implement the functions and methods described above in relation to Figures 1 to 7.

[0120] The memory 930 and storage device 960 may include various forms of volatile or non-volatile storage media. For example, the memory may include ROM (read-only memory) 931 and RAM (random access memory) 932. In this embodiment, the memory 930 may be located inside or outside the processor 910, and the memory 930 may be connected to the processor 910 via various known means.

[0121] In some embodiments, at least some of the configurations or functions of the crowd estimation-based robot delivery time and robot delivery route prediction device and method according to the embodiment can be embodied in a program or software executed on a computing device 900, and the program or software can be stored on a computer-readable medium.

[0122] In some embodiments, at least some of the configurations or functions of the crowd estimation-based robot delivery time and robot delivery path prediction device and method according to the embodiment may be embodied using the hardware or circuitry of the computing device 900, or by separate hardware or circuitry that can be electrically connected to the computing device 900.

[0123] Although embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto. Various modifications and improvements made by persons with ordinary skill in the art to which the present invention belongs, utilizing the basic concepts of the present invention as defined in the following claims, also fall within the scope of the present invention. [Explanation of Symbols]

[0124] 100: Delivery time and route prediction device based on crowd estimation 110: Prediction Department 120: Sensor Interface 130: Data Analysis Department 140: Robot Service Interface 150: Learning Department

Claims

1. The stage of setting up multiple possible routes to reach the destination; A step of collecting crowd estimation information for the plurality of routes through sensors along the plurality of routes; A step of calculating traffic and latency for each of the multiple routes based on the crowd estimation information; A step of predicting the delivery time for each of the multiple routes to the destination based on the traffic and latency; and A method for predicting delivery time and delivery route based on crowd estimation, which includes a step of determining the route with the shortest delivery time among the aforementioned multiple routes as the delivery route.

2. The step of setting up multiple routes that can be taken to the aforementioned destination is: Upon receiving a delivery request, the system includes the step of receiving a saved travel path plan from the robot and determining multiple routes to the destination where the difference in travel time between routes does not exceed a critical level, based on the said travel path plan. A method for predicting delivery time and delivery route based on crowd estimation as described in claim 1.

3. The sensors include camera sensors (including CCTV), lidar sensors, thermal imaging sensors, elevator sensors, robot sensors, and entrance / exit gate sensors. A method for predicting delivery time and delivery route based on crowd estimation as described in claim 1.

4. The step of collecting crowd estimation information for the multiple routes through sensors along the multiple routes is: This includes the step of collecting people counting information along the plurality of paths through the camera sensor and the thermal image sensor, respectively. A method for predicting delivery time and delivery route based on crowd estimation as described in claim 3.

5. The step of collecting crowd estimation information for the multiple routes through sensors along the multiple routes is: The step includes collecting the density of a specific section, including the elevator section and the entrance / exit gate section, on the route based on the entrance / exit log data obtained from the elevator sensor and the entrance / exit gate sensor. A method for predicting delivery time and delivery route based on crowd estimation as described in claim 4.

6. The step of calculating traffic and latency for each of the multiple routes based on the crowd estimation information is as follows: This includes the step of calculating the traffic based on the people counting information and the density of the specific section, and storing it in a database. A method for predicting delivery time and delivery route based on crowd estimation as described in claim 5.

7. The step of calculating traffic and latency for each of the multiple routes based on the crowd estimation information is as follows: The step includes calculating the latency based on the time between the time when the crowd estimation information is collected and the time when the traffic is calculated using the crowd estimation information. A method for predicting delivery time and delivery route based on crowd estimation as described in claim 1.

8. The step of predicting the delivery time for each of the multiple routes to the destination based on the traffic and latency is as follows: This includes a step of predicting the delivery time by combining the required time calculated based on the traffic for each of the aforementioned multiple routes and the delay time calculated based on the latency, A method for predicting delivery time and delivery route based on crowd estimation as described in claim 1.

9. After delivery is completed, the predicted delivery time and the actual delivery time for the final determined delivery route are compared, and if the comparison result satisfies certain criteria, the delivery time prediction model is updated based on the comparison result. If the comparison results do not satisfy the specified criteria, the step further includes saving the comparison results to a database. A method for predicting delivery time and delivery route based on crowd estimation as described in claim 1.

10. If the calculation of the traffic and latency is impossible due to a malfunction or error in the aforementioned sensor, The step of predicting the delivery time for each of the multiple routes to the destination based on the traffic and latency is as follows: The process further includes predicting the average delivery time based on recent traffic information, including density and people counting information for a specific period of time, stored in the aforementioned database, and recent average latency information. The step of determining the route with the shortest delivery time among the aforementioned multiple routes as the delivery route is: A step of generating multiple delivery routes based on the average delivery time and setting priorities; and The process includes determining the initial delivery route using the first route with the highest priority, and then finalizing the delivery route using a second route with the next highest priority based on traffic and latency information updated during delivery. A method for predicting delivery time and delivery route based on crowd estimation as described in claim 6.

11. A data analysis unit that collects crowd estimation information for multiple routes to the destination via sensors; A route analysis unit that calculates traffic and latency for multiple routes to the destination based on the crowd estimation information; A delivery time prediction unit that predicts the delivery time for each of the multiple routes to the destination based on the traffic and latency; and A delivery time and delivery route prediction device based on crowd estimation, including a delivery route determination unit that determines the route with the shortest delivery time among the aforementioned multiple routes as the delivery route.

12. When a delivery request is received, the data analysis unit receives the saved travel path plan from the robot and, based on the travel path plan, identifies multiple routes to the destination where the difference in travel time between routes does not exceed a critical level. A delivery time and delivery route prediction device based on crowd estimation according to claim 11.

13. The sensors include camera sensors (including CCTV), lidar sensors, thermal imaging sensors, elevator sensors, robot sensors, and entrance / exit gate sensors. A delivery time and delivery route prediction device based on crowd estimation according to claim 11.

14. The data analysis unit collects people counting information along the multiple paths through the camera sensor and the thermal image sensor, respectively. A delivery time and delivery route prediction device based on crowd estimation according to claim 13.

15. The data analysis unit collects the density of a specific section, including the elevator section and the entrance / exit gate section, on the route based on the entrance / exit log data acquired from the elevator sensor and the entrance / exit gate sensor. A delivery time and delivery route prediction device based on crowd estimation according to claim 14.

16. The route analysis unit calculates the traffic based on the people counting information and the density of the specific section, and stores it in a database. A delivery time and delivery route prediction device based on crowd estimation according to claim 15.

17. The route analysis unit calculates the latency based on the time between the time when the crowd estimation information is collected and the time when the traffic is calculated using the crowd estimation information. A delivery time and delivery route prediction device based on crowd estimation according to claim 11.

18. The delivery time prediction unit predicts the delivery time by combining the required time calculated based on the traffic for each of the multiple routes and the delay time calculated based on the latency. A delivery time and delivery route prediction device based on crowd estimation according to claim 11.

19. The system further includes a learning unit that, after delivery is completed, compares the predicted delivery time with the actual delivery time for the final determined delivery route, updates the delivery time prediction model based on the comparison result if the comparison result satisfies a specific criterion, and saves the comparison result to a database if the comparison result does not satisfy the specific criterion. A delivery time and delivery route prediction device based on crowd estimation according to claim 11.

20. If the calculation of the traffic and latency is impossible due to a malfunction or error in the aforementioned sensor, The delivery time prediction unit predicts the average delivery time based on recent traffic information, including density and people counting information for a specific period recently stored in the database, and recent average latency information. The delivery route determination unit generates multiple delivery routes based on the predicted average delivery time and sets a priority order. The first delivery route is determined using the route with the highest priority, and the final delivery route is determined using the second route with the next highest priority, based on traffic and latency information updated during delivery. A delivery time and delivery route prediction device based on crowd estimation according to claim 11.

Images