A delivery system and method for a delivery robot

Abstract

The invention relates to a delivery system and a delivery method for a semi-autonomous or autonomous delivery robot (18) tasked with delivering an order (12) corresponding to a delivery command (14) transmitted over a cloud environment (16) from an electronic platform (11). A route planner (26) is activated over the cloud environment (16) in accordance with the real-time status data (22) and the delivery command (14) is set to randomly weight possible delivery routes (30) according to a set of parameters (28). The cumulative data analysis (34) is adjusted in line with the delivery command (14), which is taken according to the delivery data (32) provided in the cloud environment (16). The randomly weighted parameter set (28) is adjusted to be weighted according to the data analysis (34) with a machine learning architecture (36).

Description

[0001] A DELIVERY SYSTEM AND METHOD FOR A DELIVERY ROBOT

[0002] TECHNICAL FIELD

[0003] The invention relates to a delivery system and a delivery method for a delivery robot in which the route is optimized.

[0004] PRIOR ART

[0005] With the expansion of the application areas of the A* algorithm (A star algorithm) in the literature and the continuous development of artificial intelligence architecture, semi- autonomous or autonomous delivery robot technology has become increasingly widespread. As a result, it has become utilized in many areas such as smart industry, military, medical care, and services. At the same time, the tasks faced by delivery robots have become increasingly complex, and the environment in which they are used has transformed from a single deterministic environment to an uncertain one. Therefore, recent research on semi- autonomous or autonomous intelligent control technology for delivery robots in complex environments has attracted intense interest from academia and industry. The methodology for determining the arrival route of robots to target points in delivery systems has also become one of the current research points in intelligent robotics as one of the key technologies.

[0006] CN110096055 invention is related to a smart food distribution navigation method and navigation system. The invention describes a delivery robot for food distribution where route planning is done in an indoor environment using the A* algorithm to shorten the distance.

[0007] BRIEF DESCRIPTION OF THE INVENTION

[0008] The object of the invention is to optimize the route for the delivery robot to the target point where the order needs to be delivered, and to minimize the cost incurred during delivery.

[0009] To achieve the mentioned objectives, the invention explains a delivery system for a semi- autonomous or autonomous delivery robot tasked with delivering an order sent from an electronic platform via a cloud environment; the system includes a controller (20) located on the delivery robot, which enables the robot's movement and facilitates data flow with the cloud environment. The real-time status data of the delivery robot is transmitted and stored via the cloud environment, which includes a central server providing order infrastructure. In the delivery system for the delivery robot, a route planner activated via the cloud environment according to real-time status data and delivery order adjusts to randomly weight possible delivery routes according to a parameter set, and it is set to perform cumulative data analysis with the delivery order received in accordance with the delivery data provided in the cloud environment. The randomly weighted parameter set is adjusted to be weighted according to data analysis with machine learning architectures. In this way, a cost is calculated for possible routes between the robot's starting and target points, and the least costly route is chosen.

[0010] In a preferred configuration of the invention, the parameter set includes at least one parameter such as the distances to the delivery address in the robot's possible delivery routes, the type of roads the robot will travel on possible delivery routes, the average arrival times achieved in order deliveries in the delivery data recorded in the cloud environment on possible delivery routes, past energy consumption data recorded in the cloud environment during the robot's movement on possible delivery routes, the time the order was placed, estimated crowd situations in the robot's possible delivery routes, road inclines in the robot's possible delivery routes, internet access strength in the robot's possible delivery routes, the accuracy of the global satellite navigation system in the robot's possible delivery routes, the proportion of drivable area where the robot moves, the shaking ratio of the robot during movement, estimated weather conditions in the robot's possible delivery routes, and the cost of the robot returning after the delivery of the order. Thus, the least costly route for the delivery robot is determined by taking into account different parameters within the parameter set.

[0011] In a preferred configuration of the invention, the route planner is an A* algorithm. In this way, the calculation of routes according to the parameter set is ensured with the A* algorithm, and the most ideal route can be given as an output.

[0012] In a preferred configuration of the invention, the delivery data provided with the delivery of the order by the delivery robot is set to be kept in the cloud environment. In this way, the most ideal route for the delivery robot to the target point is determined according to the analyses of all past order data.

[0013] In a preferred configuration of the invention, the real-time status data of the delivery robot is the battery charge percentage and location of the robot during the order period. In this way, the most ideal route is determined by including the data coming from the robot in the route calculation.

[0014] A preferred embodiment of the invention includes a connection unit that provides internet access to the cloud environment for the robot, connected to the controller located in the delivery robot. In this way, the delivery robot can facilitate data flow with the cloud environment.

[0015] A preferred application of the invention includes the steps of providing the parameter set to the route planner; receiving the delivery order related to the delivery order made to the delivery robot; the delivery robot going to the loading area for loading upon receipt of the delivery order; placing the delivery package into the delivery robot; the route planner randomly weighting possible delivery routes; determining the approximately cost-effective route and transmitting it to the delivery robot via the cloud environment, and the robot starting to move; transmitting the real-time status data of the delivery robot to the cloud environment during the order delivery; performing cumulative data analysis with previous orders; the randomly weighted possible routes being weighted in suitable operations with the data analysis performed with machine learning architectures during the order delivery of the delivery robot; and the order being delivered with the approximate arrival of the delivery robot at the delivery address. In this way, the most ideal route is obtained through the route planning algorithm and machine learning architectures with the most optimal parameters and coefficients in accordance with the method developed for the delivery system.

[0016] BRIEF DESCRIPTION OF THE FIGURES

[0017] Figure 1 is a schematic representation of a delivery system for a delivery robot.

[0018] Figure 2 is a representation of a flow chart related to a delivery method for a delivery robot.

[0019] DETAILED DESCRIPTION OF THE INVENTION

[0020] This detailed description of the invention is explained with references to examples, without any limitation on the subject matter of the invention, solely for better understanding.

[0021] Figure 1 schematically shows a delivery system for a delivery robot. In a delivery system (10) for a delivery robot, orders (12) are given from an electronic platform (11 ). The electronic platform (11 ) could be an order site accessed from a computer or mobile device, or it could also be an order platform accessed via a mobile application. For example, a delivery order (14) is formed on the electronic platform (11 ) in response to an order (12) given by a customer. The delivery order (14) is transmitted to a delivery robot (18) via a cloud environment (16). Here, the delivery robot (18) is a semi-autonomous or autonomous mobile robot responsible for delivering the received orders (12). Also, in the delivery robot, there are one or more wheels that enable forward and backward movement along the horizontal axis. In the delivery robot of the invention (18), there are four wheels. There is a suspension corresponding to each wheel that reduces vibration against obstacles such as bumps, curbs that the delivery robot (18) encounters during the delivery, and provides grip by overcoming obstacles. Suspensions are installed independently of each other. There is a controller (20) on the delivery robot (18) that allows the vehicle to work, i.e., provides the command control of the robot to enable robot movement. Additionally, the controller (20) is an electronic circuit structure that provides data flow between the delivery robot (18) and the cloud environment (16). Here, the real-time status data (22) transmitted from the controller (20) to the cloud environment (16) during the delivery of the order (12) by the delivery robot (18) is transferred to a central server (24) via the cloud environment (16). This data (22) is stored in the central server (24). Also, the central server (24) is the order infrastructure provider. The real-time status data (22) of the delivery robot (18) is the charge percentage (64) of the robot's battery and its location (66). Thus, the most ideal route (37) can be determined by including the data (22) from the delivery robot (18) in the route calculation. A route planner (26) is actively working in determining the delivery route (37) of the delivery robot (18). Here, the route planner (26) is activated over the cloud environment (16) in accordance with the real-time status data (22) and the delivery order (14). Also, the route planner (26) is a unit that randomly weights possible delivery routes (30) according to a set of parameters (28) and determines a route (37) for the delivery robot (18). The route planner (26) used in determining the route (37) is an A-star algorithm. Thus, the calculation of the routes (30) according to the parameter set is ensured, and the most ideal route (37) can be obtained as output. In the delivery system (10), a cumulative data analysis (34) is also performed in accordance with the delivery data (32) provided in the cloud environment (16) with the received delivery order (14). According to the result of the data analysis (34), by weighting the randomly weighted parameter set (28) with machine learning architectures (36), it ensures the optimization of the delivery routes (30) determined in the route planner (26) of the delivery robot (18), thereby providing the most ideal route (37). Thus, by calculating a cost for possible routes (30) between the starting points just before and the target points, and by comparing these costs, it is possible to find the most optimal route for the delivery robot (18). The route planner (26) and machine learning architectures (36) utilize factors such as road conditions, traffic data, pedestrian density, and weather conditions for optimizing the delivery routes (30).

[0022] Figure 2 shows the detailed components of the delivery robot (18) from figure 1 . The delivery robot (18) is equipped with various sensors (38), including but not limited to lidar sensors, radar sensors, ultrasonic sensors, infrared sensors, and camera sensors. These sensors (38) are used to perceive the environment and navigate the robot to its destination without any human intervention. The sensors (38) are connected to a controller (20), which processes the input from the sensors (38) and controls the robot's motion accordingly.

[0023] Figure 3 shows a schematic view of the route planner (26) in more detail. The route planner (26) calculates a set of possible routes (30) based on real-time status data (22) and the delivery order (14). The set of possible routes (30) is calculated using an A-star algorithm which is known to find the least-cost path from a given initial point to one or more goal points. This algorithm is used to find the shortest route that the delivery robot (18) needs to follow to complete the delivery. The parameters (28) used by the route planner (26) include, but are not limited to, distance, time, traffic, and road conditions. Based on these parameters (28), the route planner (26) assigns a weight to each possible route (30), and the route with the lowest total weight is selected as the most ideal route (37).

[0024] Figure 4 shows the machine learning architectures (36) used to optimize the routes determined by the route planner (26). The machine learning architectures (36) analyze the cumulative data (34) and use the result of the analysis to adjust the weights of the parameters (28) used by the route planner (26). This is done by training the machine learning model on the cumulative data (34), allowing the model to learn from the past delivery data (32) and improve its future predictions.

[0025] Figure 5 shows an example of a user interface of the electronic platform (11) where the customers can place their orders (12). The electronic platform (11 ) shows the real-time status data (22) of the delivery robot (18) including its location (66) and the battery percentage (64). It also provides an estimated delivery time based on the most ideal route (37) determined by the route planner (26) and machine learning architectures (36).

[0026] Overall, the invention provides a delivery system for a delivery robot that can optimize the delivery route based on real-time status data and cumulative data analysis. By leveraging machine learning and efficient path-finding algorithms, the delivery system can provide a faster and more efficient delivery service.

[0027] REFERENCE NUMBERS

[0028] 10 Delivery system 54 GNSS accuracy

[0029] 11 Electronic platform 56 Drivable area ratio

[0030] 12 Order 58 Robot vibration rate

[0031] 14 Delivery command 60 Estimated weather conditions

[0032] 16 Cloud environment 62 Cost of the robot's return

[0033] 18 Delivery robot 64 Battery charge percentage Controller 66 Robot's location Real-time status data 68 Connection unit Central server 70 Providing the parameter set Route planner 72 Arrival of the delivery order Parameter set 74 Going to the loading area Delivery routes 76 Placing the delivery package Delivery data 78 Random weighting Data analysis 80 Determining the route Machine learning architectures 82 Transmitting the route information Ideal route 84 The robot starts moving Delivery distance 86 Order delivery process Types of roads 88 Transmission to the cloud environment Average arrival times 90 Previous orders Energy consumption data 92 Performing data analysis Order time 94 Weighting of data analyses Estimated crowd situations 96 Arrival at the delivery address Road inclinations 98 Execution of order delivery Internet access strength

Claims

CLAIMS1- A delivery system for a delivery robot comprising a semi-autonomous or autonomous delivery robot (18) responsible for delivering an order (12) provided from an electronic platform (11 ), which is transmitted via a cloud environment (16); a controller (20) that enables the movement of the robot and provides data flow with the cloud environment (16), located on the delivery robot (18); and a central server (24) that provides order infrastructure, where the real-time status data (22) of the delivery robot (18) are transmitted and stored via the cloud environment (16) characterized in that a route planner (26) is activated via the cloud environment (16) in accordance with real-time status data (22) and the delivery command (14), and randomly weights possible delivery routes (30) according to a parameter set (28); system is set to perform a cumulative data analysis (34) with the delivery command (14), which is taken in accordance with the delivery data (32) provided in the cloud environment (16), and to weight the randomly weighted parameter set (28) in accordance with data analysis (34) utilizing a machine learning architecture (36).2- A delivery system for a delivery robot according to claim 1 , wherein at least one parameter in the parameter set (28) representing the distances to the delivery address (38) on the robot's (18) possible delivery routes (30), types of roads (40) on which the robot (18) will move on possible delivery routes (30), average arrival times (42) recorded in the delivery data (32) in the cloud environment (16) for order deliveries on the robot's (18) possible delivery routes (30), past energy consumption data (44) recorded in the cloud environment (16) during the robot's (18) movement on possible delivery routes (30), order placement time (46), estimated crowd situations (48) on the robot's (18) possible delivery routes (30), road inclinations (50) on the robot's (18) possible delivery routes (30), internet access strength (52) on the robot's (18) possible delivery routes (30), GNSS accuracy (54) of the robot's (18) possible delivery routes (30), drivable area ratio (56), robot vibration rate (58) during movement, estimated weather conditions (60) on the robot's (18) possible delivery routes (30), and the return cost (62) after the delivery (12) and outputted one more time.3- A delivery system for a delivery robot according to claim 1 , wherein a controller (20) with a built-in battery charge level monitor providing real-time battery charge percentage (64) data and a built-in GPS or GNSS module to provide the robot's location (66) data in real-time.4- A delivery system for a delivery robot according to claim 1 , wherein the controller (20) is having a connection unit (68) that provides a connection to the cloud environment (16) fortransmitting the real-time status data (22), receiving delivery command (14), and providing the parameter set (28).5- A delivery system for a delivery robot according to claim 1 , wherein the order delivery process (86) is initiated with providing the parameter set (70), upon the arrival of the delivery order (72), to the cloud environment (16), the controller (20) then commands the robot (18) to go to the loading area (74) and place the delivery package (76) after the command (14) is received, then, the route planner (26) begins to weight possible delivery routes (30) with random weighting (78) using the parameter set (28), then determining the route (80) and transmitting the route information (82) to the robot (18) after weighting, after which the robot (18) starts moving (84).6- A delivery system for a delivery robot according to claim 5, wherein during the robot's movement (84), the controller (20) continuously transmits real-time status data (22) to the cloud environment (16).7- A delivery system for a delivery robot according to claim 1 , wherein the system utilizes data (32) of previous orders (90) stored in the cloud environment (16) for performing data analysis (92).8- A delivery system for a delivery robot according to claim 1 , wherein a machine learning architecture (36) for the weighting of data analyses (94) is utilized.9- A delivery system for a delivery robot according to claim 1 , wherein upon arrival at the delivery address (96), the robot (18) executes the order delivery (98), after which the controller (20) sends a completion signal to the central server (24) via the cloud environment (16).10- A delivery system for a delivery robot according to claim 1 , wherein the system continually updates the parameter set (28) based on the real-time status data (22), past order data (90), and cumulative data analysis (34).

Images