Method to control vehicle fleets to deliver on-demand transportation services

• US 10,192,448 B2
• Filed: 09/30/2016
• Issued: 01/29/2019
• Est. Priority Date: 09/30/2016
• Status: Expired due to Fees

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Moreira-Matias, Luis, Saadallah, Amal, Khiari, Jihed
• Primary Examiner(s): Brushaber, Frederick M
• Application Number:
  US15/281,142
• Publication Number:
  US 20180096606A1
• Time in Patent Office: 851 Days
• CPC Class Codes: G05D 1/0088 characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours using knowledge based models G06N5/00
• CPC Class Codes: G05D 1/0291 Fleet control monitoring fleets in traffic control systems for road vehicles G08G1/127, G08G1/127
• CPC Class Codes: G06Q 50/40 Business processes related to the transportation industry shipping G06Q10/083
• CPC Class Codes: G07B 15/00 Arrangements or apparatus for collecting fares, tolls or entrance fees at one or more control points handling coins or valuable papers, e.g. banknotes, G07D; apparatus for vending or hiring articles or services activated by coins, credit cards, paper currency or the like G07F7/00, G07F17/00
• CPC Class Codes: G07C 5/08 Registering or indicating performance data other than driving, working, idle, or waiting time, with or without registering driving, working, idle or waiting time
• CPC Class Codes: G08G 1/202 Dispatching vehicles on the basis of a location, e.g. taxi dispatching
• CPC Class Codes: G08G 1/205 Indicating the location of the monitored vehicles as destination, e.g. accidents, stolen, rental

Abstract

A method for providing dispatching services for an on-demand transportation (ODT) service includes determining that a predictive assignment message should be transmitted to a vehicle, generating, in response to the determining that a predictive assignment should be transmitted to a vehicle, the predictive assignment message, and transmitting, to the vehicle, the predictive assignment message. Generating the predictive assignment message uses one or more prediction models computed from historical and real-time ODT service data.

Claims

1. A method for automatically distributing vehicles capable of performing on-demand transportation (ODT) services, the method comprising:

determining that a predictive assignment message should be transmitted to a vehicle;

generating, in response to the determining that a predictive assignment should be transmitted to a vehicle, the predictive assignment message; and

transmitting, to the vehicle, the predictive assignment message,

wherein generating the predictive assignment message is performed using at least two prediction models,

wherein each of the at least two prediction models is generated based, at least in part, on historical ODT service data,

wherein the at least two prediction models include one or more time-varying Poisson models and one or more time-series analysis models.

2. The method of claim 1, further comprising:

receiving, from the vehicle, a response to the predictive assignment message,

wherein the response to the predictive assignment messages includes one of an acceptance of the predictive assignment or a denial of the predictive assignment.

3. The method of claim 1, further comprising:

receiving, from a user, a request for an ODT service; and

assigning, based at least in part on a location of the user and a location included in the predictive assignment message, the requested ODT service to the vehicle.

4. The method of claim 1, further comprising:

receiving, from a plurality of users during a current time interval, a plurality of requests for ODT services; and

updating, according to the plurality of requests for ODT services, the real-time ODT service data.

5. The method of claim 4, wherein the plurality of requests for ODT services each includes a starting location for the requested ODT service.

6. The method of claim 1, wherein determining that a predictive assignment message should be transmitted to a vehicle comprises at least one of:

determining that a drop-off message has been received from the vehicle, or

determining that a predetermined duration of time has elapsed since the latter of a previous predictive assignment message has been transmitted to the vehicle or a pick-up message has been received from the vehicle.

7. The method of claim 6, wherein the drop-off message includes one or more of a timestamped GPS location, a vehicle fuel level, or a vehicle battery life.

8. The method of claim 1, wherein the one or more time-varying Poisson models include a weighted Poisson model.

9. The method of claim 8, wherein the one or more time-series analysis models include at least one of an autoregressive integrated moving average (ARIMA) model or a vector autoregression (VAR) models.

10. The method of claim 1, wherein generating the predictive assignment message using at least two prediction models comprises:

selecting, from the one or more time-varying Poisson models, a first time-varying Poisson model,

selecting, from the one or more time-series analysis models, a first time-series analysis model,

constructing an ensemble of prediction models that includes the first time-varying Poisson model and the first time-series analysis model, and

generating, using the ensemble of prediction models, the predictive assignment message.

11. The method of claim 10, wherein constructing the ensemble of prediction models includes weighting a contribution of the first time-varying Poisson model to the ensemble and weighting a contribution of the first time-series analysis model to the ensemble.

12. The method according to claim 11, wherein weighting the contribution of the first time-varying Poisson model to the ensemble and weighting the contribution of the first time-series analysis model to the ensemble is performed according to a prediction error of the first time-varying Poisson model over a number of previous periods of time and according to a prediction error of the first time-series analysis model over the number of previous periods of time.

13. The method of claim 10, wherein selecting the first time-varying Poisson model includes:

determining, for each respective time-varying Poisson model of the one or more time-varying Poisson models, a difference between an observed demand for ODT services and a demand predicted by the respective time-varying Poisson model, and

selecting, as the first time-varying Poisson model, a respective time-varying Poisson model having a smallest difference between the observed demand for ODT services and the demand predicted thereby, and

wherein selecting the first time-series analysis model includes;

determining, for each respective time-series analysis model of the one or more time-series analysis models, a difference between an observed demand for ODT services and a demand predicted by the respective time-series analysis model, and

selecting, as the first time-series analysis model, a respective time-series analysis model having a smallest difference between the observed demand for ODT services and the demand predicted thereby.

14. The method of claim 1, wherein the vehicle is an autonomously driven vehicle.

15. The method according to claim 1, wherein the historical ODT service data is historical ODT service data for a set of ODT service stands.

16. The method according to claim 15, wherein members of the set of ODT service stands are a target ODT service stand and one or more additional ODT service stands.

17. The method according to claim 16, wherein the one or more additional ODT service stands are service stands having demand correlations with the target service stand that exceed a correlation threshold.

18. The method according to claim 17, wherein a Hoeffding alarm is triggered to indicate a significant change in correlation structure between the members of the set of ODT service stands, and

wherein the members of the set of ODT service stands are reselected in response to the Hoeffding alarm being triggered.

19. A system for automatically distributing vehicles capable of performing on-demand transportation (ODT) services, the system comprising:

a central ODT dispatcher server operable to;

determine that a predictive assignment message should be transmitted to a vehicle;

generate, in response to the determining that a predictive assignment should be transmitted to a vehicle, the predictive assignment message; and

transmit, to the vehicle, the predictive assignment message,

wherein generating the predictive assignment message is performed using at least two prediction models,

wherein each of the at least two prediction models is generated based, at least in part, on historical ODT service data,

wherein the at least two prediction models include one or more time-varying Poisson models and one or more time-series analysis models.

20. A non-transitory computer readable medium having stored thereon a set of processor executable instructions for executing a method for automatically distributing vehicles capable of performing on-demand transportation (ODT) services, the processor executable instructions comprising instructions for:

determining that a predictive assignment message should be transmitted to a vehicle;

generating, in response to the determining that a predictive assignment should be transmitted to a vehicle, the predictive assignment message; and

transmitting, to the vehicle, the predictive assignment message,

wherein generating the predictive assignment message is performed using at least two prediction models,

wherein each of the at least two prediction models is generated based, at least in part, on historical ODT service data,

wherein the at least two prediction models include one or more time-varying Poisson models and one or more time-series analysis models.