Systems and methods for dynamically generating patrol schedules based on historic demand data

• US 10,558,936 B2
• Filed: 10/06/2016
• Issued: 02/11/2020
• Est. Priority Date: 10/06/2015
• Status: Active Grant

• Current Assignee: Numerica
• Original Assignee: Numerica
• Inventors: Herman, Shawn Michael, Johnson, Jason Kyle, Poore, Jeffery Brian, Du Toit, Philip Christopher
• Primary Examiner(s): Jeanty, Romain
• Application Number:
  US15/287,682
• Publication Number:
  US 20170098181A1
• Time in Patent Office: 1,223 Days
• Field of Search: 705 716
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06Q 10/06311 Scheduling, planning or task assignment for a person or group

Abstract

Systems and methods are described herein for dynamically generating and updating a patrol schedule for a shift based on historic demand event data, and a predicted-demand model based on the historic demand event data. The systems and methods also receive information associated with patrol officers assigned to a shift and constraints on the officers'"'"' availability, and generate a patrol schedule comprising patrol assignments for the assigned patrol officers optimized based on at least one policing objective. The patrol schedule may be dynamically updated based on changing information and provided to the patrol officer via mobile device or display in the patrol officer'"'"'s vehicle.

Claims

1. A system for dynamically generating and updating a patrol schedule for a shift, the system comprising:

one or more storage mediums storing executable instructions; and

one or more processors configured to execute the instructions, wherein

execution of the instructions causes the system to perform a method comprising;

receiving historic demand event data comprising information representing at least one prior incident for which patrol personnel were requested, such information representing at least one prior incident comprising, for each at least one prior incident, a location type and an incident type,

generating, by a processor of the one or more processors, a predicted-demand model configured to predict when and where future incidents will take place and how many patrol personnel will be needed to respond to each future incident by;

analyzing, by a machine learning algorithm, the historic demand event data;

generating, based on the analysis, one or more feature vectors; and

identifying a correlation between a feature vector and an incidence rate,

receiving information associated with at least one patrol officer assigned to the shift, wherein the information associated with the at least one patrol officer comprises at least one constraint on the at least one patrol officer'"'"'s availability,

generating, by the processor, a patrol schedule for a shift based on (i) the historic demand event data, (ii) the predicted-demand model, and (iii) the information associated with the at least one patrol officer assigned to the shift, wherein the patrol schedule comprises at least one patrol assignment for the at least one patrol officer assigned to the shift,

optimizing the patrol schedule based on at least one policing objective and on the at least one constraint,

outputting the optimized patrol schedule to at least one device for use by the at least one patrol officer assigned to the shift for patrolling a sector, wherein the optimized patrol schedule comprises, for each patrol officer, a sequence of geographic regions to be visited and a corresponding sequence of durations that specify how much time should be spent patrolling each geographic region,

receiving real-time information for the at least one police officers on the shift,

monitoring the received real-time information for a trigger to initiate a patrol shift schedule update;

automatically updating the optimized patrol schedule based on the real-time information, wherein updating the optimized patrol schedule comprises;

generating a second patrol schedule for a shift based on (i) the historic demand event data, (ii) the predicted-demand model, and (iii) the real-time information associated with the at least one patrol officer; and

optimizing the second patrol schedule based on the at least one policing objective and the at least one constraint, and

transmitting, to a device associated with an officer, the updated patrol schedule.

2. The system of claim 1, wherein the patrol schedule is optimized based multiple policing objectives and the multiple policing objectives are weighted to reflect relative importance.

3. The system of claim 1, wherein the method further comprises updating the patrol schedule based on a change in at least one of the at least one constraints.

4. The system of claim 3, wherein the change in the at least one constraint comprises a change in availability of one or more patrol officers.

5. The system of claim 3, wherein the change in the at least one constraint comprises a request from a patrol officer for a different patrol assignment.

6. The system of claim 1, wherein the predicted-demand model generated based on the historic demand event data is specific to a particular geographic area and is generated based on historic demand event data for that particular geographic area.

7. The system of claim 6, wherein the geographic area is partitioned into patrol regions, the patrol regions having been determined based on road information, the historic demand event data, and neighborhood information.

8. The system of claim 7, wherein the geographic area is partitioned into two or more patrol regions, the patrol regions determined automatically based on road information, the historic demand event data, and neighborhood information and such that at least one objective is satisfied, wherein the at least one objective is chosen from the group comprising:

1) the boundary lengths between regions are minimized,

     2) the sizes of the regions are balanced in terms of number of historical incidents within each region; and

     3) each region is road-connected such that any point of a region can be reached by vehicle on roads.

9. The system of claim 1, wherein generating the patrol schedule for the shift further comprises generating the patrol schedule based at least in part on correlative data related to the historic demand event data.

10. The system of claim 1, wherein outputting the patrol schedule for use by the at least one patrol officer comprises outputting the at least one patrol assignment for a patrol officer to a mobile display that travels with the patrol officer while on patrol.

11. The system of claim 10, wherein the system further outputs actionable information related to the region of assigned patrol, persons of interest in the region, addresses of frequent criminal activity, or types of incidents likely to occur in the region.

12. The system of claim 11, wherein the actionable information displayed is determined based on the location of the patrol officer or the patrol officer'"'"'s vehicle, as determined by at least one location acquisition system.

13. The system of claim 1, wherein the information associated with the at least one patrol officer assigned to the shift comprises an indication of interruptibility, the indication representing the likelihood that the patrol officer will be reassigned upon occurrence of a new incident.

14. The system of claim 1, wherein the method further comprises updating the patrol schedule periodically.