System and method for infrastructure-based parking

Abstract

An infrastructure-based autonomous parking garage solution 1000 that manages the parking of multiple vehicles simultaneously. The system is equipped with a plurality of cameras 2210, 2220, 2230, 2240

Description

BACKGROUND

[001] In modern urban environments, parking has become an increasingly complex issue for drivers and passengers. The increasing density of traffic flow and the limited availability of parking spaces have made finding a suitable spot for parking a vehicle a daunting task. Notably, the problem becomes even more pronounced near popular destinations where the demand for parking spaces is high.

[002] In addition to the difficulty in finding available parking lots, drivers also struggle to stay updated on the immediate availability of parking spaces. The current systems do not provide real-time information about vacant parking spaces, resulting in wasted time and fuel as drivers search for available spots. Another significant issue is the prohibitive cost of parking, especially in city centers and popular destinations, which can deter people from using their vehicles.

[003] Moreover, remembering the exact location where the vehicle was parked can be challenging, particularly in large parking structures. The lack of a reliable system to guide drivers back to their parked vehicles often leads to unnecessary stress and wasted time. This problem is further compounded in multi-level parking garages where navigating through the different levels can be confusing for drivers.

[004] Several attempts have been made to address these issues, but they have been met with limited success. Some solutions have focused on providing real-time information about parking space availability, but these systems often fail to provide accurate and reliable data. Other solutions have attempted to reduce parking costs by offering discounted rates during off-peak hours, but these initiatives have not substantially alleviated the overall parking cost burden for drivers.

[005] Furthermore, existing parking management systems have not adequately addressed the issue of vehicle location memory. While some systems offer a basic feature to mark the location of the parked vehicle, they lack the capability to provide precise navigation back to the vehicle, especially in large and complex parking structures.

[006] Therefore, there is a need to overcome the problems discussed above by introducing an autonomous parking management system. SUMMARY

[007] The primary objective of the present disclosure is to provide an infrastructurebased autonomous parking garage software solution that can take over control of the vehicles that go inside and plan and command them to park. This is achieved through the deployment of advanced sensors such as bird-eye view RGB cameras (e.g., with night vision capabilities and radars), and a centralized computer with abundant graphica and parallel computing capabilities.

[008] Another objective of the present disclosure is to optimize the overall time and energy used for all vehicles spent on parking by causing less traffic, plan the paths of cars more adequately, and drivers don't need to find a spot by themselves driving around. This is achieved by planning all parking maneuvers and sequence for every vehicle inside the garage altogether, based on, e.g., a ranking, graph sort algorithm or deep learning.

[009] Yet another objective of the present disclosure is to save a lot of resources and money by eliminating the need to put sets of expensive sensors and software for autonomous parking on each car. This is achieved by having, e.g., a centralized perception planning and control system, where each vehicle does not need to trigger their own Automatic Parking Assist(APA) software or camera / lidar sensors to park.

[0010] Still another objective of the present disclosure is to optimize the overall macro parking behaviors / maneuvers within a parking garage. This is achieved by, e.g., monitoring the situation of all the vehicles inside the garage, receiving data from each vehicle, performing path planning or motion planning of all the vehicles in the garage as a whole and controlling each vehicle.

[0011] According to one aspect of the present disclosure, the autonomous parking management system comprises a communication system (e.g., WiFi communication interface) configured, when deployed in a parking garage having a plurality of cameras and a plurality of radar sensors, to establish communication with the plurality of cameras and the plurality of radar sensors, and to establish communication with multiple vehicles entering the parking garage or in the parking garage. The system also includes one or more computing devices communicatively coupled to the communication system, and configured to receive camera sensor information from the plurality of cameras, and radar sensor information from the plurality of radar sensors; receive multiple respective parking request messages from the multiple vehicles, wherein the multiple respective parking request messages are for requesting parking within the parking garage; perform a planning operation that identifies multiple parking spaces in the parking garage to assign to the multiple vehicles; determine respective multiple paths for the multiple vehicles to follow to reach the multiple parking spaces identified; generate respective multiple sets of vehicle control commands based on the camera sensor information and the radar sensor information, wherein each of the multiple sets of vehicle control commands are for controlling acceleration, steering, and braking of a respective vehicle to maneuver to a respective parking space assigned to the respective vehicle; cause the communication system to communicate or transmit the respective multiple sets of vehicle control commands to the multiple vehicles.

[0012] According to another aspect of the present disclosure, the one or more computing devices are configured to identify the multiple parking spaces to assign to the multiple vehicles based on their respective vehicle types (e.g., sedan versus SUV), vehicle models, or vehicle sizes. The one or more computing devices are also configured to determine the respective vehicle sizes of the multiple vehicles based on the radar sensor information. The one or more computing devices are further configured to determine, for at least one of the multiple respective parking request messages, a predicted parking duration for which a respective vehicle associated with the at least one parking request message is estimated to park in the parking garage, and wherein a respective parking space assigned to the respective vehicle is identified by the one or more computing devices based on the predicted parking duration.

[0013] According to yet another aspect of the present disclosure, the one or more computing devices are configured to determine, for each of the multiple vehicles, a parking priority level of the vehicle relative to the other ones of the multiple vehicles, and wherein the one or more computing devices are configured, when the parking garage includes a first set of parking spaces closer to a parking garage exit relative to a second set of parking spaces, to identify the multiple parking spaces based on a planning algorithm that, for each of the multiple vehicles, increases the vehicle’s probability of being assigned to the first set of parking spaces instead of the second set of parking spaces as the vehicle’s parking priority level increases relative to the other ones of the multiple vehicles.

[0014] According to still another aspect of the present disclosure, the one or more computing devices are configured to track, based on the camera sensor information and / or the radar sensor information, respective positions of the multiple vehicles in the parking garage; track, based on the camera sensor information and / or radar information, respective positions of obstacles (e.g., pillars in the parking garage) between the multiple vehicles and their assigned parking spaces, and wherein the multiple sets of vehicle control commands are generated based on the respective positions of the multiple vehicles and based on the respective positions of the obstacles.

[0015] The proposed system offers several advantages over traditional parking solutions. It minimizes the overall time and energy used for all vehicles spent on parking by causing less traffic and planning the paths of cars more adequately. It also saves a lot of resources and money by eliminating the need to put sets of expensive sensors and software for autonomous parking on each car. Furthermore, it optimizes the overall macro parking behaviors / maneuvers within a parking garage by monitoring the situation of all the vehicles inside the garage, receiving data from each vehicle, planning as a whole and controlling each vehicle. The foregoing paragraphs have been provided by way of general introduction and are not intended to limit the scope of the following claims. The described embodiments, together with further advantages, will be best understood by reference to the following detailed description taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1: Figure 1 depicts a parking management system, a parking garage, cameras and radar sensors in the parking garage, and vehicles in the parking garage, according to embodiments herein.

[0017] Figure 2: Figure 2 is a block diagram of a parking management system, according to embodiments herein.

[0018] Figure 3: Figure 3 provides a flow diagram illustrating a method performed by a parking management system, according to embodiments herein. DETAILED DESCRIPTION

[0019] Aspects of the present disclosure are best understood by reference to the description set forth herein. All the aspects described herein will be better appreciated and understood when considered in conjunction with the following descriptions. It should be understood, however, that the following descriptions, while indicating preferred aspects and numerous specific details thereof, are given by way of illustration only and should not be treated as limitations. Changes and modifications may be made within the scope herein without departing from the spirit and scope thereof, and the present disclosure herein includes all such modifications.

[0020] The proposed system of the present disclosure is an innovative infrastructurebased autonomous parking garage software solution that can take over control of the vehicles that go inside the parking garage and command them to park. In an embodiment, the parking garage is equipped with bird-eye view RGB cameras with night vision capabilities and radars all on the cell facing down vertically, each covering a wide ground area (for example, 10m x 10m). The camera is for video and image capturing during the day, night vision is for the night, both can be used to detect vehicles, pedestrians, and obstacles such as pillars in the parking space.

[0021] In an embodiment, the radars installed on the ceiling of the parking garage provide depth information which assists the central computer to determine vehicles' and objects' heights and ground information for better planning. For instance, the system can park low cars under low ceilings. The system also has communication devices such as radio or Bluetooth to communicate with the cars. Furthermore, the system has a centralized computer with abundant graphical and parallel computing capabilities to calculate planning and control maneuvers for multiple vehicles at once.

[0022] When a vehicle enters the maneuver area of the parking garage, or wishes to leave the parking garage, a two-way Vehicle-to-infrastructure (V2I) communication between the vehicle and garage is established. The garage system takes over the control (steering, pedal and brake) of the vehicle, via either radio or wireless Internet, or Bluetooth. The central computer then proceeds to calculate a proper parking spot for the vehicle to park in, based on current parking space availability, the vehicle's car model, parking time and other status of the garage such as the congestion situation and traffic flow inside the garage. Once it finds a valid spot and plans a viable path for the vehicle to drive and park in / park out, it sends out the corresponding control commands to control the vehicle to park in the space that is designated.

[0023] This parking garage can monitor the situation of all the vehicles inside the garage, receive data from each vehicle, plan as a whole and control each vehicle. In this case, the garage can maintain the parking situation as a whole, control all the vehicles, plan and schedule each car's movement. This kind of management has a lot of benefits compared to onboard autonomous parking solutions.

[0024] The system can minimize the overall time and energy used for all vehicles spent on parking by causing less traffic, planning the paths of cars more adequately, and drivers don't need to find a spot by themselves driving around. In comparison, traditional autonomous parking is done by each individual vehicle, so each car will care for its own best maneuver and spots so there is no overall planning intelligence. For example, with the camera and radar on the ceiling, we can detect the size and type of each vehicle and park them in corresponding parking spaces like compact or larger parking spots to minimize wasted space and achieve optimal parking capacity.

[0025] With a centralized perception planning and control system, each vehicle does not need to trigger their own APA (automatic parking assist) software or camera / lidar sensors to park so there is no need to put sets of expensive sensors and software for autonomous parking on each car. Therefore, cars don't need to carry all these heavy payloads around, and thus save a lot of resources and money.

[0026] The overall macro parking behaviors / maneuvers within a parking garage can be optimized. We can guarantee parking behavior for all cars, so space can be adequately used without some drivers parking their cars wrongly to take up too much space. We can guarantee the parking performance of vehicles, without worrying that some OEMs have better parking software than others.

[0027] The proposed system is new in that it does not focus on parking one single vehicle, but all vehicles’ overall global behavior planning considering that parking maneuvers of some vehicles will affect the behaviors and decisions of other, especially neighboring vehicles. Also, this system does not require any sensor installation / heavy computation device on the vehicle, because the sensors and computers were installed on the infrastructure. Only communication modules between each vehicle and the infrastructure central brain are needed. In this case the garage infrastructure functions as an “edge computing device” where the computation takes place near the car peripheral but not on the car itself.

[0028] The planning algorithm not only will plan and control one vehicle to park, but more importantly, it plans all parking maneuvers and sequence for every vehicles inside the garage altogether, based on a ranking, graph sort algorithm or deep learning. This large-scale planning can ensure that the cars that are trying to park in the garage in one point of time can be accommodated based on their time of entry and their respective positions with other vehicles inside the garage. This would be more useful, especially during peak hours.

[0029] The prior systems may focus on methodology for a vehicle to percept using its own sensors and plan itself inside a parking garage where GNSS signal is unreliable or lost. Meanwhile this inventive system focuses on providing perception and planning from the infra only. In summary, the described system not only addresses the immediate needs for detailed parking monitoring and management but also provides a scalable and flexible solution that can evolve with technological advancements and changing user requirements. The embodiment of the disclosure described here is intended to provide a thorough understanding of the disclosure's capabilities and should not be considered limiting of its scope.

[0030] FIG. 1 provides an illustrative representation of an autonomous parking management system 1000 in a parking garage 2000 equipped with multiple cameras 2210 / 2220 / 2230 / 2240 (also illustrated in FIG. 2) and radar sensors 2110 / 2120 / 2130. In an embodiment, the system 1000 includes a communication system 1100 and one or more computing devices 1200. The cameras and radar sensors may be strategically positioned across the parking garage to capture comprehensive data about the vehicles 3100 / 3200 / 3300 and their surroundings. The communication system 1100 establishes and maintains a two-way Vehicle-to-lnfrastructure (V2I) communication with the vehicles entering or existing the parking garage, enabling the system 1000 to take control of the vehicles' movement within the parking garage.

[0031] In an embodiment, the cameras 2210 / 2220 / 2230 / 2240 in the parking garage 2000 are equipped with bird-eye view RGB capabilities and night vision features to capture video and image data during day and night. These cameras may be particularly useful for detecting the presence and movement of vehicles, pedestrians, and obstacles within the parking space. The radar sensors 2110 / 2120 / 2130, may be particularly useful for obtaining depth information, which assists a central computer or other computing device 1200 in determining the height of vehicles and objects, and the ground information for optimal parking planning. For instance, low-height cars can be parked under low ceilings to maximize space usage.

[0032] In an embodiment, the communication system 1100 of the parking management system 1000 is configured to establish communication with the cameras 2210 / 2220 / 2230 / 2240 and radar sensors 2110 / 2120 / 2130 and the vehicles 3100 / 3200 / 3300 entering or in the parking garage 2000. This communication system can communicate via, e.g., wireless internet, WiFi, or Bluetooth®, ensuring seamless data transmission between the vehicles and the computing computer 1200. The computing device 1200 is configured to receive, via the communication system, camera sensor information from the cameras and radar sensor information from the radar sensors via the communication system 1100.

[0033] In an embodiment, vehicles entering the parking garage or already inside the parking garage may request that the parking management system 1000 guide these vehicles to respective parking spots. More specifically, the vehicles may transmit or otherwise communicate parking request messages to the parking management system 1000. Upon receiving a parking request message from a vehicle, the computing system 1200 calculates a suitable parking spot for the vehicle based on, e.g., current parking space availability, the vehicle model, parking time, and the garage's congestion situation and traffic flow. It may then perform path planning to determine a path for the vehicle to reach the parking spot, and then generates a set of vehicle control commands to guide the vehicle to its designated parking spot.

[0034] In an embodiment, the computing device 1200 of the parking management system 1000 also has the capability to determine multiple parking spots for multiple vehicles to simultaneously park in the parking garage, and may control the multiple vehicles simultaneously, ensuring efficient utilization of parking spaces and smooth traffic flow within the garage. In an embodiment, it can automatically guide a vehicle to the exit area of the parking garage for easy pick-up by the customer, thereby eliminating the need for the customer to walk into the garage or look for their vehicle.

[0035] In an embodiment, the autonomous parking management system 1000 offers a centralized perception, planning, and control system, eliminating the need for each vehicle to have its own set of sensors or automatic parking assist (APA) software. This can result in significant cost savings as the system reduces the need for expensive sensors and software on each car for autonomous parking. Furthermore, the system 1000 helps minimize the overall time and energy used for parking by planning the paths of cars more adequately and reducing traffic within the parking garage.

[0036] The detailed description above provides an understanding of the overall structure and functioning of the autonomous parking management system 1000 as depicted in FIG. 1. However, it should be noted that the system can be modified or enhanced without departing from the scope of the disclosure. For example, the system can be configured to work with different types of sensors or communication protocols, or it can be integrated with other systems or services to offer additional features or capabilities.

[0037] Referring to FIG. 2, a block diagram of an autonomous parking management system is presented, designated generally by reference numeral 1000. The parking management system 1000 is designed to automate the process of parking vehicles in a parking garage, utilizing sensor data from a suite of cameras 2210, 2220, 2230, and 2240 and radar sensors (not shown in FIG. 2) for perception. The system 1000 may include a set of computing devices 1200 for planning and control, and a communication module 1100 that facilitates two-way communication between the system 1000 and the vehicles 3100, 3200, and 3300. The computing devices 1200 are responsible for processing data from the cameras and radar sensors, planning parking maneuvers, and generating vehicle control commands. As illustrated in FIG. 2, the parking management system 1000 may have a non-transitory computer-readable medium 1300, such as computer memory, that stores various modules which may be executed by the computing devices 1200. For instance, the computer memory or other non-transitory computer-readable medium may include a camera fusion module 1310, an overall scheduling and planning module 1320, an individual path planning module 1330, and a vehicle motion control module 1340, which are discussed below in more detail.

[0038] In an embodiment, the cameras 2210, 2220, 2230, and 2240 may be strategically positioned throughout the parking garage to provide comprehensive coverage of the parking area. These cameras are capable of capturing video and still images during both day and night, thanks to their night vision capabilities. The cameras can detect vehicles, pedestrians, and obstacles such as pillars within the parking area. The radar sensors may provide depth information, helping the system 1000 determine the heights of vehicles and objects, and ground information for planning purposes. The parking management system 1000 may receive image information from the cameras (e.g., via the communication system / module 1100), and process the image information via the camera fusion module 1310 to detect presence of other vehicles, pedestrians, and obstacles in the parking garage.

[0039] In an embodiment, the communication module 1100 establishes a two-way vehicle-to-infrastructure (V2I) communication with the vehicles 3100, 3200, and 3300 that enter the parking garage or wish to leave it. This communication is established via radio, wireless Internet, or Bluetooth, allowing for the exchange of information and control commands.

[0040] In an embodiment, the computing devices 1200 have ample graphical and parallel computing capabilities, enabling them to identify vehicles and / or obstacles in the parking garage, and to perform calculations for path planning and motion planning to control maneuvers for multiple vehicles to park at once. When a vehicle enters the parking garage and requests assistance from the parking management system 1000 to park in the garage, the computing devices 1200 may calculate a suitable parking spot for the vehicle based on current parking space availability, the vehicle's model, parking time, and other factors such as the congestion situation and traffic flow inside the garage. In an embodiment, the suitable parking spot may be calculated based on sensor data from the cameras and radar sensors. The computing devices 1200 may execute the overall scheduling and planning module 1320 to identify, e.g., which vehicle(s) has priority in terms of access to parking spots or right-of-way over other vehicles when moving in the parking garage, and may execute the individual path planning module 1330 to determine individual paths for respective vehicles to follow to reach their assigned parking spots.

[0041] Once the computing devices 1200 have identified a valid parking spot and planned a viable path for the vehicle, they generate the corresponding control commands (e.g., generated by the vehicle motion control module 1340) which may be executed by the vehicle to follow the path and reach the parking spot. These commands, which control, e.g., the vehicle's steering, acceleration, and braking, are then transmitted to the vehicle via the communication module 1100. In this way, the parking management system 1000 can automatically park the vehicle in the designated space. Similarly, when a vehicle needs to leave the parking garage, the system 1000 can automatically bring the vehicle to a designated pick-up area, such as the exit area of the parking garage.

[0042] The parking management system 1000 does not just handle the parking of individual vehicles. It also monitors the overall situation in the parking garage, receiving data from each vehicle and planning as a whole. This holistic approach allows the system 1000 to manage the parking situation comprehensively, controlling all vehicles and scheduling each car's movement. This is a significant advantage over traditional autonomous parking solutions, which are focused on parking individual vehicles without considering the overall parking situation.

[0043] Thus, the parking management system 1000 can plan and control the parking maneuvers and sequence for every vehicle in the garage based on, e.g., a ranking, graph sort algorithm, or deep learning. This large-scale planning ensures that all vehicles trying to park in the garage at one point can be accommodated efficiently, especially during peak hours. Moreover, the system 1000 reduces the need for sensor installation or heavy computation device on the vehicle. All the necessary sensors and computers are installed on the parking garage infrastructure, and only communication modules between each vehicle and the infrastructure computing devices 1200 of the system 1000 are needed. The garage infrastructure thus functions as an “edge computing device” where the computation takes place near the car peripheral but not on the car itself.

[0044] FIG. 3 illustrates a flow diagram that demonstrates the method conducted by the computing system 1200 of the autonomous parking management system 1000, as per the embodiments of the present disclosure. The process starts with step 4100, where the autonomous parking management system (1000) establishes communication with multiple vehicles entering or already inside the parking garage (2000). The communication is established via a communication system (1100), which is also connected to various cameras (2210 / 2220 / 2230 / 2240) and radar sensors (2110 / 2120 / 2130) installed in the parking garage. These cameras and radar sensors are responsible for capturing real-time data about the vehicles and their surroundings, including vehicle dimensions, positions, and the availability of parking spaces.

[0045] In step 4200, the system receives multiple parking request messages from the vehicles. These messages are requests from the vehicles for a parking space within the parking garage. Each parking request message carries specific information about the vehicle, such as its size, model, and expected parking duration. This information is used by the system to make informed decisions in the subsequent steps.

[0046] Step 4300 involves the system performing a planning operation. The planning operation is carried out by one or more computing devices (1200) that are communicatively coupled to the communication system (1100). The task of these computing devices is to process the data received from the cameras and radar sensors, and to use this information to identify suitable parking spaces for the vehicles. The identification of parking spaces is based on various factors, including the availability of parking spaces, the dimensions of the vehicles, and the expected parking duration.

[0047] In step 4400, the system determines multiple paths for the vehicles to reach their assigned parking spaces. This involves calculating the optimal route for each vehicle to reach its designated parking space, taking into account the current positions of the vehicles, the locations of the parking spaces, and the presence of any obstacles in the parking garage.

[0048] Step 4500 involves the generation of multiple sets of vehicle control commands. These commands are generated based on the sensor information received from the cameras and radar sensors. Each set of vehicle control commands is designed to control the acceleration, steering, and braking of a respective vehicle, enabling it to maneuver to its assigned parking space. The control commands are designed to ensure safe and efficient navigation of the vehicles within the parking garage.

[0049] Finally, in step 4600, the system communicates the vehicle control commands to the vehicles. This is done via the communication system (1100), which transmits the commands to the vehicles using radio or wireless internet, or Bluetooth. Once the vehicles receive the commands, they can autonomously navigate to their assigned parking spaces, guided by the control commands.

[0050] This method, as illustrated in FIG. 3, enables effective management of parking within a parking garage. By using an infrastructure-based autonomous parking management system, the process of parking is simplified and made more efficient. The system eliminates the need for drivers to manually find a parking space, reducing the time and energy spent on parking. Furthermore, by planning the paths of vehicles and controlling their movements, the system can minimize traffic within the parking garage, further enhancing the efficiency of the parking process.

[0051] The present disclosure introduces an infrastructure-based autonomous parking management system. This advanced system is designed to control the movement of multiple vehicles entering a parking garage, and plan and command them to park. The parking garage is equipped with bird-eye view RGB cameras with night vision capabilities and radar sensors on the cell facing down vertically. Each sensor covers a wide ground area, for example, a 10m x 10m zone.

[0052] The cameras serve a dual purpose. During the day, they function as video and image capturing devices. At night, their night vision capabilities come into play. These cameras are used to detect vehicles, pedestrians, and obstacles such as pillars in the parking space. The radar sensors provide depth information, which assists the central computer in determining vehicles' and objects' heights and ground information for better planning. For instance, the system can park low cars under low ceilings.

[0053] The system also incorporates communication devices such as radio or Bluetooth to communicate with the vehicles. The system is further equipped with a centralized computer with abundant graphical and parallel computing capabilities. This computer is used to calculate planning and control maneuvers for multiple vehicles at once.

[0054] When a vehicle enters the maneuver area of the parking garage, or wishes to leave the parking garage, a two-way V2I (vehicle-to-infrastructure) communication between the vehicle and garage is established. The garage system takes over the control of the vehicle, via either radio or wireless Internet, or Bluetooth. The central computer then proceeds to calculate a suitable parking spot for the vehicle to park in, based on factors such as current parking space availability, the vehicle's model, parking time, and other conditions of the garage such as the congestion situation and traffic flow.

[0055] Once it finds a valid spot and plans a viable path for the vehicle to drive and park in / park out, it sends out the corresponding control commands to control the vehicle to park in the space that is designated. With the same control and planning capability, the parking garage can also automatically bring the car to the exit area of the parking garage. This allows the customer to pick up their vehicles with ease, without walking into the garage or looking for their vehicles.

[0056] This parking garage can monitor the situation of all the vehicles inside the garage, receive data from each vehicle, plan as a whole and control each vehicle. This holistic approach to parking management has several benefits compared to onboard autonomous parking solutions. It can minimize the overall time and energy used for all vehicles spent on parking by causing less traffic, planning the paths of cars more adequately, and drivers don't need to find a spot by themselves driving around.

[0057] Furthermore, with a centralized perception planning and control system, each vehicle does not need to trigger their own APA (automatic parking assist) software or camera / lidar sensors to park. Therefore, cars don't need to carry all these heavy payloads around, and thus save a lot of resources and money. The overall macro parking behaviors / maneuvers within a parking garage can also be optimized. With the proposed system, space can be adequately used without some drivers parking their cars wrongly to take up too much space.

[0058] The parking management system of the present disclosure is new in that it does not focus on parking one single vehicle, but all vehicles’ overall global behavior planning. It considers that parking maneuvers of some vehicles will affect the behaviors and decisions of other, especially neighboring vehicles. Also, this system does not require any sensor installation / heavy computation device on the vehicle, because the sensors and computers were installed on the infrastructure. Only communication modules between each vehicle and the infrastructure central brain are needed.

[0059] More specifically, the planning algorithm not only plans and controls one vehicle to park, but more importantly, it plans all parking maneuvers and sequence for every vehicle inside the garage altogether, based on a ranking, graph sort algorithm, or deep learning. This large-scale planning can ensure that the cars that are trying to park in the garage at one point of time can be accommodated based on their time of entry and their respective positions with other vehicles inside the garage. This would be more useful, especially during peak hours.

[0060] The system's innovative approach to parking management presents several advantages over traditional methods. It provides a centralized perception, planning, and control system, eliminating the need for each vehicle to have its own set of sensors or automatic parking assist (APA) software. This results in cost savings as the system reduces the need for expensive sensors and software for autonomous parking on each car. Furthermore, the system helps minimize the overall time and energy used for parking by planning the paths of cars more adequately and reducing traffic within the parking garage.

[0061] In addition, the system can monitor the overall situation in the parking garage, receiving data from each vehicle and planning as a whole. This holistic approach allows the system to manage the parking situation comprehensively, controlling all vehicles and scheduling each car's movement. This is a significant advantage over traditional autonomous parking solutions, which are focused on parking individual vehicles without considering the overall parking situation.

[0062] These embodiments are only illustrative of the inventive concepts contained herein. Other embodiments and modifications may be made to the compositions and methods without departing from the spirit and scope of the disclosure. Therefore, the scope of the present disclosure should not be limited to the embodiments described herein but should be defined by the appended claims and their equivalents.

Claims

1. An autonomous parking management system (1000), comprising:a communication system (1100) configured, when deployed in a parking garage (2000) having a plurality of cameras (2210 / 2220 / 2230 / 2240) and a plurality of radar sensors (2110 / 2120 / 2130), to establish communication with the plurality of cameras (2210 / 2220 / 2230 / 2240) and the plurality of radar sensors (2110 / 2120 / 2130), and to establish communication with multiple vehicles (3100 / 3200 / 3300) entering the parking garage (2000) or in the parking garage (2000);one or more computing devices (1200) communicatively coupled to the communication system 1100, and configured to: (i) receive (4100), via the communication system, camera sensor information from the plurality of cameras, and radar sensor information from the plurality of radar sensors; (ii) receive (4200), via the communication system, multiple respective parking request messages from the multiple vehicles (3100 / 3200 / 3300), wherein the multiple respective parking request messages are for requesting parking within the parking garage; (iii) perform a planning operation (4300) that identifies multiple parking spaces in the parking garage to assign to the multiple vehicles (3100 / 3200 / 3300); (iv) determine (4400) respective multiple paths for the multiple vehicles to follow to reach the multiple parking spaces identified in step (4300); (v) generate (4500) respective multiple sets of vehicle control commands based on the camera sensor information and the radar sensor information, wherein each of the multiple sets of vehicle control commands are for controlling acceleration, steering, and braking of a respective vehicle to maneuver to a respective parking space assigned to the respective vehicle; (vi) cause (4600) the communication system (1100) to communicate or transmit the respective multiple sets of vehicle control commands to the multiple vehicles (3100 / 3200 / 3300).

2. The autonomous parking management system of claim 1, wherein the one or more computing devices are configured to identify the multiple parking spaces to assign to the multiple vehicles based on their respective vehicle types, vehicle models, or vehicle sizes.

3. The autonomous parking management system of claim 2, wherein the one or more computing devices are configured to determine the respective vehicle sizes of the multiple vehicles based on the radar sensor information.

4. The autonomous parking management system of claim 1, wherein the one or more computing devices are configured to determine, for at least one of the multiple respective parking request messages, a predicted parking duration for which a respective vehicle associated with the at least one parking request message is estimated to park in the parking garage, and wherein a respective parking space assigned to the respective vehicle is identified by the one or more computing devices based on the predicted parking duration.

5. The autonomous parking management system of claim 4, wherein the one or more computing devices are configured, when the parking garage has multiple parking levels, to identify the multiple parking spaces to assign to the multiple vehicles based on a determination of, for each of the multiple vehicles, which parking level to assign to the respective vehicle, wherein the parking level for the vehicle is determined based on the predicted parking duration of the respective vehicle.

6. The autonomous parking management system of claim 1, wherein the one or more computing devices are configured to determine, for each of the multiple vehicles, a parking priority level of the vehicle relative to the other ones of the multiple vehicles, and wherein the one or more computing devices are configured, when the parking garage includes a first set of parking spaces closer to a parking garage exit relative to a second set of parking spaces, to identify the multiple parking spaces based on a planning algorithm that, for each of the multiple vehicles, increases the vehicle’s probability of being assigned to the first set of parking spaces instead of the second set of parkingspaces as the vehicle’s parking priority level increases relative to the other ones of the multiple vehicles.

7. The autonomous parking management system of claim 6, wherein the one or more computing devices are configured to determine respective parking priority levels for the multiple vehicles based on their respective garage entry times.

8. The autonomous parking management system of claim 1, wherein the one or more computing devices are configured to track, based on the camera sensor information and / or the radar sensor information, respective positions of the multiple vehicles in the parking garage, track, based on the camera sensor information and / or radar information, respective positions of obstacles between the multiple vehicles and their assigned parking spaces, and wherein the multiple sets of vehicle control commands are generated based on the respective positions of the multiple vehicles and based on the respective positions of the obstacles.

9. The autonomous parking management system of claim 1, wherein the one or more computing devices are configured to determine, based on the radar sensor information, respective vehicle heights of the multiple vehicles, wherein the multiple parking spaces are identified based on the respective vehicle heights of the multiple vehicles.

10. The autonomous parking management system of claim 1, wherein the one or more computing devices are configured to identify multiple parking spaces or to determine the respective multiple paths for the multiple vehicles based on traffic congestion in the parking garage or traffic flow inside the parking garage.

11. A method performed by one or more computing devices (1200) of an autonomous parking management system (1000), the method comprising: (i) receiving (4100), via the communication system, camera sensor information from aplurality of cameras (2210 / 2220 / 2230 / 2240) in the parking garage (2000), and radar sensor information from a plurality of radar sensors (2110 / 2120 / 2130) in the parking garage (2000);(ii) receiving (4200), via a communication system (1100) of the autonomous parking management system (1000), multiple respective parking request messages from multiple vehicles entering a parking garage (2000) or in the parking garage (2000);(iii) identifying (4300) multiple parking spaces in the parking garage (2000) to assign to the multiple vehicles;(iv) determining (4400) respective multiple paths for the multiple vehicles to follow to reach the multiple parking spaces, wherein each of the multiple sets of vehicle control commands are for controlling acceleration, steering, and braking of a respective vehicle to maneuver to a respective parking space assigned to the respective vehicle;(v) generating (4500) respective multiple sets of vehicle control commands based on the camera sensor information and the radar sensor information;(vi) causing (4600) the communication system (1100) to communicate or transmit the respective multiple sets of vehicle control commands to the multiple vehicles.

12. The method of claim 11, wherein the multiple parking spaces are identified based on their respective vehicle types, vehicle models, or vehicle sizes.

13. The method of claim 12, wherein the respective vehicle sizes of the multiple vehicles are determined based on the radar sensor information.

14. The method of claim 11, further comprising determining, for at least one of the multiple respective parking request messages, a predicted parking duration for which a respective vehicle associated with the at least one parking request message is estimated to park in the parking garage, wherein respective parking spaces for the respective vehicle is identified based on the predicted parking duration.

15. The method of claim 14, wherein the multiple parking spaces are identified based on a determination of, for each of the multiple vehicles, which parking level to assign to the respective vehicle, wherein the parking level for the vehicle is determined based on the predicted parking duration of the respective vehicle.

16. The method of claim 11, wherein a parking priority level of each of the multiple vehicles is determined relative to the other ones of the multiple vehicles, and wherein the multiple parking spaces are identified based on a planning algorithm that, for each of the multiple vehicles, increases the vehicle’s probability of being assigned to a first set of parking spaces closer to a parking garage exit relative to a second set of parking spaces as the vehicle’s parking priority level increases relative to the other ones of the multiple vehicles.

17. The method of claim 16, wherein the respective parking priority levels for the multiple vehicles are determined based on their respective garage entry times.

18. The method of claim 11, wherein the multiple sets of vehicle control commands are generated based on respective positions of the multiple vehicles in the parking garage, and based on respective positions of obstacles between the multiple vehicles and their assigned parking spaces.

19. The method of claim 11, wherein the multiple parking spaces are identified based on respective vehicle heights of the multiple vehicles, which are determined based on the radar sensor information.

20. The method of claim 11, wherein the multiple parking spaces or the respective multiple paths for the multiple vehicles are determined based on traffic congestion in the parking garage or traffic flow inside the parking garage.

Images