Patent Issued for Predicting parking availability (USPTO 11928964): United Services Automobile Association

2024 MAR 28 (NewsRx) -- By a News Reporter-Staff News Editor at Insurance Daily News -- A patent by the inventors Dama, Nikhil (Apex, NC, US), Evans, Laura Frisbie (Roswell, GA, US), Haberland, Christopher Ryan (Seattle, WA, US), Lu, Ting (San Antonio, TX, US), Stolarczyk, Simon Phillip (San Antonio, TX, US), Williams, Kristopher Thomas (Helotes, TX, US), filed on April 4, 2022, was published online on March 12, 2024, according to news reporting originating from Alexandria, Virginia, by NewsRx correspondents.

Patent number 11928964 is assigned to United Services Automobile Association (San Antonio, Texas, United States).

The following quote was obtained by the news editors from the background information supplied by the inventors: “Finding parking in a busy location can be very frustrating. The frustration increases when a driver’s particular needs, such as for handicapped parking within 30 feet of a Chinese restaurant, reduce the amount of suitable parking. When a suitable space is occupied with no other vacant suitable spaces around, the driver may choose to wait until the space opens up or may instead choose to drive somewhere else in hopes of finding a vacancy. However, facing this choice only adds to the driver’s frustration because most drivers cannot call on any information for making an informed choice of how best to find suitable parking in the shortest amount of time.

“The techniques introduced here may be better understood by referring to the following Detailed Description in conjunction with the accompanying drawings, in which like reference numerals indicate identical or functionally similar elements.”

In addition to the background information obtained for this patent, NewsRx journalists also obtained the inventors’ summary information for this patent: “Aspects of the present disclosure are directed to a parking availability predictor system for providing a prediction of parking availability that meets a user’s specified needs. When the user submits a query for parking availability, he can include desired parking criteria such as location, handicapped status, and desired time to begin parking. The parking availability predictor system consults a parking occupancy model which gathers current and historical information about parking occupancy, including parking durations and durations of vacancies between occupancies. The model processes the gathered information into a statistical representation of probable parking occupancy and vacancy. The parking availability predictor system uses the occupancy model to predict when parking that meets the user’s needs will become available. That prediction, including suitable parking locations along with the expected start of their vacancies, is then sent to the user in answer to his query.

“The parking availability predictor system can be accessed when a user submits a parking availability query and receives the prediction using any known technology. In some variations, an application running on the user’s personal computer or smart phone allows the user to select search criteria and associated importance ratings, possibly storing these in a profile for repeated use. In addition, a user can specify absolute filtering criteria that can be applied before any predictions are returned to the user. As one example, a handicapped driver or driver with a handicapped passenger may allow handicapped parking to be returned in the predictions, while a prediction for a non-handicapped driver would not include handicapped parking. The application’s user interface (UI) can be graphical which allows for map-based selection of desired locations and presentation of the results. An example of this is presented below in FIG. 6 and the accompanying text. Some variations support a text or voice UI. In some variations, the returned prediction can be fed into, or the user’s query and response application can be integrated into, a navigational device in a smart phone or automobile.

“A parking occupancy model supports the parking availability predictor system. In order to support the parking availability system in providing predictions suited to each user’s search criteria, the parking occupancy model, in some variations, includes relatively static characteristics about parking places such as size, a location of the place, whether the place is for handicapped parking, whether the place is for automobile or motorcycle parking, whether the place is well lighted, whether the place is covered, whether the place requires payment, whether the place is secured, a safety rating for the place, or whether overnight parking is allowed. Such static parking characteristics can be obtained from various sources such as city planning information, parking business providing parking information for the spaces they manage, analysis of images of the parking spaces by machine learning models trained to identify parking space characteristics, manual tagging of parking spaces, crowd sourced information of users supplying parking space characteristics as the parking availability predictor system is used, etc.

“In addition to the relatively static parking characteristics listed above, the parking occupancy model processes information about parking occupancy. This information can be gathered from special purpose parking sensors, street cameras, car-tracking devices, and other sources. The parking occupancy model uses this information in at least two ways. First, the information allows the parking occupancy model to tell whether a parking place is currently occupied or vacant and for how long it has been occupied or vacant. Second, the parking occupancy model accumulates such information over time and processes it into a statistical representation that tells how long a given parking place tends to remain occupied and how long it tends to remain vacant. For example, the occupancy model can know that people parking in front of a given restaurant tend to park on average for about two hours. The representation can be trained via machine-learning or by traditional statistical techniques.

“When the parking availability predictor system receives a search query from a user, it consults the parking occupancy model. Potential parking is filtered by whatever parking characteristics the user includes in the query, the filtering in some variations informed by the importance ratings the user assigned to search criteria. The occupancy model allows the parking availability predictor system to predict when suitable parking will become available. Building on the example given above, if the occupancy model knows that people tend to park for two hours in front of a given restaurant, and if three user-suitable spaces in front of the restaurant have been occupied for almost two hours, then the occupancy model knows that suitable parking will probably become available at expected vacancy times calculated for each place as the difference between the expected parking duration (two hours) and the measured parking duration of each place’s current occupancy.”

The claims supplied by the inventors are:

“1. A method for a computer system to provide a prediction of parking availability, the method comprising: receiving a parking search query specifying search criteria including one or more characteristics for parking spaces; generating one or more imminent-vacancy scores for parking by filtering parking spaces that match the one or more characteristics and applying a model to the filtered parking spaces, wherein the model was generated based at least on: durations of parking-place occupancies for multiple parking places; and durations of parking-place vacancies for the multiple parking places; and providing, based on the one or more imminent-vacancy scores, a parking availability prediction, the providing comprising one or more of: generating a map, based on the one or more imminent-vacancy scores, of expected vacancies; providing directions to an area with the highest imminent-vacancy score of the one or more imminent-vacancy scores; or providing an expected wait time, for parking, determined based on the one or more imminent-vacancy scores.

“2. The method of claim 1 wherein the filtering is performed using parking place features, logged in the model, including one or more of: size, parking place location, parking place handicapped status, parking place lighting status, parking place covered status, parking place payment requirements, parking place safety rating, maximum parking duration, or any combination thereof.

“3. The method of claim 1 wherein the model is a machine learning model generated using training data created from the durations of parking-place occupancies for multiple parking places and the durations of parking-place vacancies for the multiple parking places.

“4. The method of claim 1, wherein the providing the parking availability prediction comprises providing how well indicated parking spaces match the search criteria.

“5. The method of claim 4 wherein the search criteria includes one or more of: when the user wishes to park, a minimum space size, a maximum distance from a location, a space handicapped parking status, a space payment requirement, a desired time to park, or any combination thereof.

“6. The method of claim 4 further comprising receiving from a user one or more importance rankings, each associated with one or more of the search criteria.

“7. The method of claim 1 wherein the parking availability prediction includes the expected wait time and the map of expected vacancies.

“8. A non-transitory computer-readable storage medium storing instructions that, when executed by a computing system, cause the computing system to perform a process for providing a prediction of parking availability, the process comprising: receiving a parking search query specifying search criteria including one or more characteristics for parking spaces; generating one or more imminent-vacancy scores for parking by filtering parking spaces that match the one or more characteristics and applying a model to the filtered parking spaces, wherein the model was generated based at least on: durations of parking-place occupancies for multiple parking places; and durations of parking-place vacancies for the multiple parking places; and providing, based on the one or more imminent-vacancy scores, a parking availability prediction, the providing comprising one or more of: generating a map, based on the one or more imminent-vacancy scores, of expected vacancies; providing directions to an area with the highest imminent-vacancy score of the one or more imminent-vacancy scores; or providing an expected wait time, for parking, determined based on the one or more imminent-vacancy scores.

“9. The non-transitory computer-readable storage medium of claim 8 wherein the filtering is performed using parking place features including one or more of: size, parking place handicapped status, parking place lighting status, parking place covered status, maximum parking duration, or any combination thereof.

“10. The non-transitory computer-readable storage medium of claim 8 wherein the statistical model generated using regression analysis from the durations of parking-place occupancies for multiple parking places and the durations of parking-place vacancies for the multiple parking places.

“11. The non-transitory computer-readable storage medium of claim 8 wherein the providing the parking availability prediction comprises providing how well indicated parking spaces match the search criteria.

“12. The non-transitory computer-readable storage medium of claim 11 wherein the search criteria includes one or more of: when the user wishes to park, a minimum space size, a maximum distance from a location, a space handicapped parking status, a space payment requirement, a desired time to park, or any combination thereof.

“13. The non-transitory computer-readable storage medium of claim 8, wherein the process further comprises receiving from a user one or more importance rankings, each associated with one or more of the search criteria.

“14. The non-transitory computer-readable storage medium of claim 8 wherein the parking availability prediction includes the expected wait time.

“15. A computing system for providing a prediction of parking availability, the computing system comprising: one or more processors; and one or more memories storing instructions that, when executed by the one or more processors, cause the computing system to perform a process comprising: receiving a parking search query specifying search criteria including one or more characteristics for parking spaces; generating one or more imminent-vacancy scores for parking by filtering parking spaces that match the one or more characteristics and applying a model to the filtered parking spaces, wherein the model was generated based at least on: durations of parking-place occupancies for multiple parking places; and durations of parking-place vacancies for the multiple parking places; and providing, based on the one or more imminent-vacancy scores, a parking availability prediction, the providing comprising one or more of: generating a map, based on the one or more imminent-vacancy scores, of expected vacancies; providing directions to an area with the highest imminent-vacancy score of the one or more imminent-vacancy scores; or providing an expected wait time, for parking, determined based on the one or more imminent-vacancy scores.

“16. The computing system of claim 15 wherein the filtering is performed using parking place features including one or more of: size, parking place handicapped status, parking place lighting status, parking place covered status, maximum parking duration, or any combination thereof.

“17. The computing system of claim 15 wherein the statistical model generated using regression analysis from the durations of parking-place occupancies for multiple parking places and the durations of parking-place vacancies for the multiple parking places.

“18. The computing system of claim 15 wherein the providing the parking availability prediction comprises providing how well indicated parking spaces match the search criteria.

“19. The computing system of claim 15, wherein the process further comprises receiving from a user one or more importance rankings, each associated with one or more of the search criteria.

“20. The computing system of claim 15 wherein the parking availability prediction includes the expected wait time.”

URL and more information on this patent, see: Dama, Nikhil. Predicting parking availability. U.S. Patent Number 11928964, filed April 4, 2022, and published online on March 12, 2024. Patent URL (for desktop use only): https://ppubs.uspto.gov/pubwebapp/external.html?q=(11928964)&db=USPAT&type=ids

(Our reports deliver fact-based news of research and discoveries from around the world.)