US 12,093,855 B2
Systems and methods for service location optimization
Asif Gulambhai Pachigar, Slough (GB); Anuj Sethi, Reading (GB); Matthew D. Scherzer, Maplewood, NJ (US); Michael Weinstock, Harrisburg, PA (US); Michael C. Pedi, Chicago, IL (US); Sumeet Pundlik, Bengaluru (IN); Kanika Goyal, Madhya Pradesh (IN); Debasis Panda, Odisha (IN); Suman Kumar Rana, Hyderabad (IN); and Pranav Alva, New York, NY (US)
Assigned to MARS, INCORPORATED, McLean, VA (US)
Filed by Mars, Incorporated, McLean, VA (US)
Filed on Sep. 20, 2022, as Appl. No. 17/933,544.
Claims priority of application No. 202211036805 (IN), filed on Jun. 27, 2022.
Prior Publication US 2023/0419201 A1, Dec. 28, 2023
Int. Cl. G06Q 10/06 (2023.01); G06Q 10/063 (2023.01); H04W 64/00 (2009.01)
CPC G06Q 10/063 (2013.01) [H04W 64/003 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method for optimizing an operational characteristic of a service location, the method comprising:
receiving past performance data for one or more operational characteristics for one or more reference service locations;
wherein the past performance data comprises customer transactions from a plurality of operational POS terminals;
receiving, from a user, a target value for an additional operational characteristic of a target service location;
receiving operational data for one or more operational point of sale (POS) terminals for the target service location and/or from the reference service locations;
training a machine learning model using the received past performance data and the received operational data for the operational POS terminals;
calculating a predicted service rate or a predicted arrival rate using the trained machine learning model;
redistributing one or more customer transactions among the operational POS terminals based on the predicted service rate or the predicted arrival rate;
calculating an average value of the additional operational characteristic for each of the operational POS terminals following the redistributing;
determining an average POS terminal as a POS terminal among the one or more operational POS terminals with the calculated average value of the additional operational characteristic that matches the target value for the additional operational characteristic;
selecting the average POS terminal as a representative POS terminal;
setting operational characteristics of a target POS terminal of the target service location to be equal to the operational characteristics of the representative POS terminal; and
receiving, from the target POS terminal, updated operational performance data representing completed transactions after the setting operational characteristics for the one or more operational characteristics; and
retraining the machine learning model using the received operation characteristics and the received updated performance data for the operational POS terminals.