Patent Application
20240240957
The disclosure relates to methods and apparatus for providing a message comprising a starting time of a predicted navigation route to a user of the vehicle.
Navigation applications in vehicles and on mobile terminal devices can inform a driver of a vehicle after the destination has been entered whether there is more traffic than usual on the planned navigation route to the destination. In addition, navigation applications can proactively inform the user of a vehicle about a journey time to a destination at the start of the journey.
It is an object of the disclosure to provide a starting time for a predicted navigation route to a user of a vehicle in an efficient manner. In particular, one object of the disclosure is to efficiently provide a starting time for a predicted navigation route to a predicted destination and for a predicted departure time to the user of the vehicle, so that the user of the vehicle can reach the predicted destination on time.
The above-stated objects, as well as others, are achieved by at least some advantageous embodiments and developments disclosed herein.
A first aspect is characterized by a method for providing a message comprising a starting time of a predicted navigation route to a user of the vehicle. The method can be a computer-implemented method. The method is preferably carried out on a vehicle-external server. Alternatively, the method can be carried out on an engine control unit or computer of the vehicle. Alternatively, the method can be carried out on a combination of the vehicle-external server and control unit or computer of the vehicle. The vehicle can be a motor vehicle. The method comprises receiving a last vehicle position of a journey of the user with the vehicle. The last vehicle position can be a vehicle position ascertained using a global satellite navigation system. Preferably, the last vehicle position is received from the vehicle by a vehicle-external server.
The method further comprises determining a current vehicle location using the received last vehicle position of the journey of the user. The current vehicle location can be a cluster of vehicle positions representing the current vehicle location. The method further comprises predicting a departure time window for the start of the next journey, a destination of the next journey, and a probability for the start of the next journey to the predicted destination in the predicted departure time window for the user of the vehicle on the basis of the determined current vehicle location. If the probability of the beginning of the next journey to the predicted destination in the predicted departure time window exceeds a specified threshold value, the method ascertains a navigation route of the user of the vehicle to the predicted destination, the method determines the estimated journey time of the ascertained navigation route of the user of the vehicle to the predicted destination, and the method determines the starting time of the predicted navigation route on the basis of the ascertained estimated journey time. Furthermore, the method transmits the message comprising the starting time of the predicted navigation route to the user of the vehicle if the starting time of the predicted navigation route corresponds to the current time.
Advantageously, the method can notify a user of the vehicle proactively about the starting time a of predicted navigation route, so that the user can reach the destination on time using the predicted navigation route, taking into account a current traffic situation, when the user begins the journey with the vehicle at the starting time. No manual settings are required for the user of the vehicle regarding the navigation destination and/or the navigation route. The navigation system of the vehicle learns the user's behavior dynamically. Furthermore, the navigation system of the vehicle can efficiently assist the user in planning journeys with the vehicle more easily and in making journeys, especially regularly recurring journeys, with the vehicle. The use of the vehicle by the user is efficiently simplified by the vehicle efficiently assisting the user in planning the starting time of a journey.
According to an exemplary embodiment, the method can further comprise, if the probability of the start of the next journey to the predicted destination in the predicted departure time window does not exceed the specified threshold value, predicting a new departure time window for the start of the next journey, a destination of the next journey, and a probability of the start of the next journey to the predicted destination in the predicted time window for the user of the vehicle on the basis of the determined current vehicle location after expiry of a first, specified time interval. This allows efficient control over the prediction of the next journey with the vehicle.
According to one or more embodiments, ascertaining the navigation route of the user of the vehicle to the predicted destination can comprise ascertaining a preferred navigation route of the user of the vehicle as the navigation route of the user, and ascertaining a fastest navigation route as the navigation route of the user, if no preferred navigation route of the user of the vehicle has been determined. This allows the navigation route to the predicted destination to be efficiently adapted to the user of the vehicle.
According to some embodiments, ascertaining the estimated journey time of the ascertained navigation route of the user of the vehicle to the predicted destination can comprise determining a starting time for ascertaining an estimated journey time of the ascertained navigation route of the user of the vehicle to the predicted destination, and ascertaining the estimated journey time of the ascertained navigation route of the user of the vehicle to the predicted destination as soon as the starting time for ascertaining the estimated journey time of the ascertained navigation route is reached. This allows efficient control over ascertaining the estimated journey time. The estimated journey time is ascertained only at the times and/or time intervals relevant to the user.
According to some embodiments, the method can further comprise terminating the method, in particular a current step of the method, if a movement of the vehicle is detected. This can effectively prevent the user from receiving messages at starting times that are out of date and/or no longer relevant to the user of the vehicle.
According or more exemplary embodiments, the method can further comprise monitoring the start of the journey after the message has been provided to the user of the vehicle. If the journey does not start after the message is provided to the user of the vehicle, the method can predict a new departure time window for the start of the next journey, a destination of the next journey, and a probability of the start of the next journey to the predicted destination in the predicted time window for the user of the vehicle on the basis of the determined current vehicle location after expiry of a second, specified time interval. This allows the method to be efficiently adapted to the user's behavior regarding the start of the journey.
According to some embodiments, the departure time window can be increased if no departure time window has been predicted, for which the probability of the start of the next journey to the predicted destination in the predicted departure time window exceeds the predetermined threshold value. This allows the departure time window to be efficiently adjusted such that the user is provided with a starting time for a relevant navigation route to a relevant, predicted destination by means of a message. The relevance of the information provided to the user can be efficiently increased.
A further aspect is characterized by a computer-readable medium for providing a message comprising a starting time of a predicted navigation route to a user of the vehicle, wherein the computer-readable medium comprises instructions that, when executed on a computer and/or a control unit, carry out the method described above.
A further aspect is characterized by a system for providing a message comprising a starting time of a predicted navigation route to a user of the vehicle, wherein the system is designed to carry out the method described above.
A further aspect is characterized by a vehicle comprising the system described above for providing a message comprising a starting time of a predicted navigation route to a user of the vehicle.
A further aspect is characterized by a mobile terminal device comprising the system described above for providing a message comprising a starting time of a predicted navigation route to a user of the vehicle.
Additional features arise from the claims, the figures and the description of the figures. All the features and feature combinations cited in the description above, and the features feature combinations cited in the description of the figures below and/or shown in the figures alone are applicable not only in the respective combination indicated, but also in other combinations or else in isolation.
In the following, an exemplary embodiment is described by reference to the attached drawings. This will reveal further details, preferred embodiments and extensions of the.
In detail,
The method 100 is preferably carried out on a user-specific basis. The starting time of a predicted navigation route is calculated and provided individually for each user of the vehicle. A user-specific message is provided to each user of the vehicle. The method 100 can receive 102 a last vehicle position of a journey of the user with the vehicle. For example, the vehicle-external server can receive 102 the last vehicle position of a journey by the user from the vehicle. The last vehicle position of the journey is transmitted from the vehicle to the external server preferably in real time, or approximately in real time. The vehicle-external server thus has a current, latest vehicle position available, so that the method 100 can also notify the user of a brief stay by the user at the last vehicle position before the start of the next journey.
The method 100 can determine 104 a current vehicle location using the received last vehicle position of the user's journey. The last vehicle position received can be a vehicle position of a global navigation satellite system, or GNSS for short. The received last vehicle position may vary slightly in terms of vehicle location. For example, e vehicle may be parked in slightly different parking spaces on a road or in a parking garage, all of which are located near a user's home or workplace. Furthermore, the GNSS vehicle position may vary due to an inaccuracy in determining the GNSS vehicle position. By determining the current vehicle location, inaccuracies of GNSS vehicle positions and/or different parking positions and/or parking locations can be efficiently taken into account. For example, the current vehicle location can be determined by calculating clusters for all past final vehicle positions. Known algorithms such as DBSCAN and/or agglomerative clustering can be used to calculate the clusters. If clusters have already been calculated for all past final vehicle positions, a center point of a cluster located nearest the last vehicle position received can be ascertained as the current vehicle location.
The method 100 can predict 106 a departure time window for the start of the next journey, a destination of the next journey, and a probability for the start of the next journey to the predicted destination in the predicted departure time window for the user of the vehicle on the basis of the determined current vehicle location. In addition or alternatively, the method 100 can predict an arrival time window. The steps described below can be applied in an analogous manner to the prediction of an arrival window. The following describes the prediction of the departure time window in detail. The departure time window for the start of the next journey and the destination can be predicted in different ways. In a first option, only the departure time window for the start of the next journey can be predicted first, followed by the probability for the destination, assuming that the next journey takes place in the predicted departure time window. The probability P for the start of the next journey to the predicted destination Z in the predicted departure time window F is obtained as follows:
The departure time window for the start of the next journey can be calculated by searching for a time window with a fixed duration, for example a time window of 30 minutes, within a specified time interval, for example within a specified time interval of 24 h from the end of the last journey, which maximizes the probability of the start of the next journey. Examples of time windows are: 5:00 p.m. to 5:30 p.m., 5:30 p.m. to 6:00 p.m., or 6:00 p.m. to 6:30 p.m. Alternatively, the fixed-duration time window can be shifted by small steps, for example 5 min, to find a departure time window with a greater departure probability. Examples of time windows are: 5:00 p.m. to 5:30 p.m., 5:05 p.m. to 5:35 p.m., or 5:10 p.m. to 5:40 p.m. A departure probability within a time window can be calculated using historical departure times of a user at the current vehicle location. The time window with the highest departure probability can be predicted as the departure time window.
The departure probability within a time window is determined on the basis of the current vehicle location ascertained, as follows:
In order to calculate P (Journey to Z| Next journey within F), the simplifying assumption can be made that the next journey will take place in the middle of the time window F or at a different time within the time window F. This has the advantage that an algorithm for destination prediction can be used. For example, the destination prediction algorithm can predict the destination on the basis of the current vehicle location and the current time. In detail, the destination of the next journey can be predicted with a classification algorithm that determines a probability for each possible destination of the vehicle. The destination can be a vehicle location, which, as described above, has been ascertained by means of a known clustering algorithm using historical vehicle positions of the user. The predicted destination can correspond to the vehicle location, or the cluster representing the vehicle location, with the highest probability. In order to predict a geo-position as the destination, a cluster center can be calculated. For example, the cluster center can be calculated using an arithmetic mean of all geo-positions of vehicle positions of the cluster.
Specifically, the following features can be used to predict the destination:
The prediction can be performed using a known statistical model and/or using a trained machine learning method. Examples of machine learning methods are: logistical regression, support vector machines, random forest, artificial neural networks, boosting, or K-Nearest Neighbor algorithm. The result of predicting the destination is a probability of each destination. The most probable destination is selected as the predicted destination.
Alternatively, the departure time window and the destination can be predicted at the same time, i.e. the probability that the start of the next journey to the predicted destination Z will take place in the predicted departure time window F is calculated directly. In other words, P (next journey to Z within F) is calculated directly. Specifically, for each possible destination Z, a time window FZ can be sought which maximizes the probability of the next journey to the destination Z starting in the time window FZ. The destination Z, for which the time window FZ has the highest probability, is predicted as the destination. The simultaneous prediction of the departure time window and the destination has the advantage that a combination of departure time window and destination with the highest probability is found. Another advantage over a calculation with (1) is that P (Journey to Z| Next journey within F) does not need to be approximated by choosing an assumed starting time. In other words, the calculation of the probability for the predicted arrival time window and the predicted destination is preferentially more accurate for large time windows, for example a time window of 2 h. Furthermore, no algorithm for destination prediction is required when the departure time window and the destination are predicted simultaneously.
If the probability of the beginning of the next journey to the predicted destination in the predicted departure time window exceeds a specified threshold value, for example a threshold value of 0.7, the method 100 uses this predicted departure time window to the predicted destination for the further steps of the method 100.
If the probability of the start of the next journey to the predicted destination in the predicted departure time window does not exceed a specified threshold value, for example a threshold value of 0.7, the prediction of the departure time window and the destination can be repeated at specified times and/or in specified time intervals. The departure time window may change over time, even if the vehicle is not moved. For example, if the vehicle was parked in the evening on the previous day and there is a possibility that another journey would take place in the evening, the probability that the next journey on the following morning would be to the destination Work may be lower.
If the prediction of the departure time window takes place in the vehicle, the vehicle would have to be woken up at a later time if the calculation was repeated. This can be avoided by the calculation of future times taking place already when the vehicle is parked, assuming that the vehicle will not be moved before the future time. If the vehicle is parked e.g. at 5 p.m., then a calculation can be made for the future times 5:10 p.m., 5:20 p.m., . . . up to a maximum time horizon of e.g. 24 h from the end of the last journey, in each case assuming that the vehicle has not been moved before this time. The calculation ends as soon as a time window is found that has a probability higher than the threshold. If, contrary to the assumption, a movement of the vehicle at an earlier time is detected, the method will be terminated as described above. This method can also be advantageous when the calculation is performed on a vehicle-external server, as this avoids re-calculations.
If no departure time window and destination have been predicted with a probability higher than the specified threshold value, the departure time window can be incrementally increased up to a specified maximum size before the departure time window and destination are predicted again. For example, the specified maximum size of the departure time window can be 2 h. If a specified initial size of the departure time window is 30 minutes, the departure time window can be increased incrementally from 30 minutes to 60 minutes, 90 minutes, and 120 minutes. The user of the vehicle may, for example, have a usual departure time for a journey from home to work in a departure time window of, for example, 30 minutes and a usual departure time from work to home in a larger departure time window of, for example, 2 hours. A larger departure time window may result in the user receiving a message about the starting time of a predicted route, in which the user will start a journey to the predicted destination with a high probability within the larger departure time window. If no departure time window and destination is predicted with a probability greater than the specified threshold value even for the maximum size of the departure time window, the prediction of the arrival time window and the destination can be repeated at a later time. In this case the departure time window can be increased again, as described above.
Further, the method 100 can ascertain a navigation route of the user of the vehicle to the predicted destination 108. The navigation route can be ascertained using previous navigation routes of the user. If the probability is above a predefined threshold value, for example 0.6 or 0.7, this navigation route is ascertained as the navigation route of the user to the predicted destination. If no navigation route could be determined using past navigation routes of the user, a fastest navigation route can be determined as the user's navigation route.
For the ascertained navigation route of the user of the vehicle to the predicted destination, the method 100 can ascertain 110 an estimated journey time. To ascertain 110 the estimated journey time, the method can retrieve data on current traffic information from a traffic information service, for example from a real-time traffic information service, RITI service for short.
Furthermore, the method 100 can calculate a starting time for ascertaining an estimated journey time. For example, the starting time for ascertaining the estimated journey time can be calculated as follows:
For a small, predicted departure time window of, for example, 30 min, the start of the predicted departure time window can be used for calculating the starting time. For example, for a large, predicted departure time window of, for example, 2 h, the center of the predicted departure time window can be used for calculating the starting time. The use of the center of the predicted departure time window for large departure time windows has the advantage that a notification is not sent to the user too early and the current traffic information has a greater relevance for the start of the journey of the vehicle user. A buffer may be required when calculating the starting time so that the user has sufficient time to reach the vehicle. The buffer can be fixed in advance. Alternatively, the buffer can be set dynamically depending on a current position of the user and/or a current position of the vehicle.
Alternatively, the starting time can also be calculated using the predicted arrival time window:
The calculation of the starting time using the predicted arrival time window may be more accurate if a variance of the arrival time at the predicted destination is less than a variance of the departure time at the predicted destination. Once the starting time for ascertaining the estimated journey time is reached, the method 100 can calculate the estimated journey time and/or query it from the traffic information service at regular intervals.
The method 100 can determine 112 a starting time of the predicted navigation route depending on the ascertained estimated journey time and transmit 114 the message comprising the starting time of the predicted navigation route to the user of the vehicle if the starting time of the predicted navigation route corresponds to a current time. Specifically, the starting time of the predicted navigation route can be calculated as follows:
Starting time of the predicted navigation route=Start or center of the predicted departure time window Ascertained estimated journey time—Buffer.
A current position of the user may differ from a current position of the vehicle. Therefore, before the message is provided to the user, it is also possible to check whether the position of a mobile terminal device of the user is close to the current position of the vehicle. If the position of the mobile terminal device is not close to the current position of the vehicle, the delivery of the message to the user can be prevented. For example, if the user has gone on vacation without the vehicle and the position of the mobile terminal is therefore not close to the current position of the vehicle, the method can prevent the user of the vehicle from receiving messages while on vacation. In addition or alternatively, the user's appointment or calendar entries can be used to prevent incorrect messages being sent to the user. If the current time corresponds to the starting time of the predicted navigation route and/or if none of the above exclusion criteria applies, the method 100 can send the user a message comprising the starting time 114. The message may include additional information about a possible alternative route. Furthermore, the message can be provided on a mobile terminal of the vehicle user and/or on a navigation system of the vehicle. The mobile terminal can be a smartphone, a smartwatch, and/or a pair of augmented reality goggles.
If the user does not start the journey after the message has been delivered, the method 100 can be carried out again after a specified time interval, for example a time interval of 3 h. If the user starts the journey before the message is delivered, the method 100 is interrupted and/or the message is prevented from being transmitted to the user.
Advantageously, the method can provide the user with departure times to destinations which are relevant to the user. Furthermore, the method can prevent the user from receiving unnecessary messages, which efficiently increases the convenience of the method. The user does not need to make any manual entries regarding the destination, navigation route and/or departure time. The user receives the information relevant to the user with regard to the predicted departure time window and the predicted destination automatically.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2021 116 780.9 | Jun 2021 | DE | national |
The present application is the U.S. national phase of PCT Application PCT/EP2022/055678 filed on Mar. 7, 2022, which claims priority of German patent application No. 102021116780.9 filed on Jun. 30, 2021, which is incorporated herein by reference in its entirety.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2022/055678 | 3/7/2022 | WO |
