Methods and apparatus for determining wait times
A system is provided and generally includes a server, an associate computing device, and a customer computing device. The server may receive data from the customer computing device indicating that a customer is picking up items from a predetermined location. The server may compute an estimated wait time for the customer based on one or more machine learning processes. In some examples, a number of unexpected customers that may arrive is determined. The machine learning process may compute the estimated wait time based on the number of unexpected customers. The machine learning process may be trained with historical data. The estimated wait time is transmitted to the customer computing device, and is displayed to the customer. In some examples, the server sends a list of customers waiting to be serviced to the associate computing device. The list may be prioritized based on estimated wait times for those customers.
1. A system comprising:
a first computing device communicatively coupled to a second computing device, wherein the first computing device is configured to:
receive first data identifying an arrival time window of a customer from the second computing device;
determine a current time based on the received first data;
determine a first number of customers waiting for service;
determine a first number of associates available to assist the first number of customers;
determine an estimated number of unexpected customers that may arrive during the arrival time window based on historical data, the historical data identifying numbers of unexpected customers during a same arrival time window for previous periods of time;
determine an expected wait time for the customer based on the current time, the first number of customers waiting for service, the estimated number of unexpected customers that may arrive during the arrival time window, and the first number of associates available; and
transmit the expected wait time to the second computing device.
2. The system of claim 1 , wherein determining the expected wait time comprises applying at least one machine learning process to the current time, the first number of customers waiting for service, and the first number of associates available.
3. The system of claim 2 , wherein the at least one machine learning process is based on XGBoost.
4. The system of claim 1 , wherein the first computing device is configured to:
determine at least one historic wait time corresponding to the arrival time window over a previous period of time;
determine a second number of customers waiting for service during the at least one historic wait time;
determine a second number of associates available during the at least one historic wait time; and
determine the expected wait time for the customer based on the at least one historic wait time, the second number of customers, and the second number of associates.
5. The system of claim 1 , wherein the second computing device is configured to display the expected wait time.
6. The system of claim 1 , wherein the second computing device is configured to:
receive the expected wait time;
determine whether the expected wait time is beyond a threshold;
display a first screen if the expected wait time is beyond the threshold; and
display a second screen if the expected wait time is not beyond the threshold.
7. The system of claim 6 , wherein the first screen identifies the expected wait time.
8. The system of claim 1 , wherein the second computing device is configured to:
receive the expected wait time;
display a screen in response to receiving the expected wait time; and
in response to displaying the screen:
receive a first input identifying a parking location for the customer; and
transmit the parking location to the first computing device.
9. The system of claim 1 , wherein the second computing device is configured to:
receive the expected wait time;
display a screen in response to receiving the expected wait time; and
in response to displaying the screen:
receive a first input identifying a vehicle color for the customer; and
transmit the vehicle color to the first computing device.
10. A method comprising:
receiving first data identifying an arrival time window of a customer from a second computing device;
determining a current time based on the received first data;
determining a first number of customers waiting for service;
determining a first number of associates available to assist the first number of customers;
determine an estimated number of unexpected customers that may arrive during the arrival time window based on historical data, the historical data identifying numbers of unexpected customers during a same arrival time window for previous periods of time;
determining an expected wait time for the customer based on the current time, the first number of customers waiting for service, the estimated number of unexpected customers that may arrive during the arrival time window, and the first number of associates available; and
transmitting the expected wait time to the second computing device.
11. The method of claim 10 , wherein determining the expected wait time comprises applying at least one machine learning process to the current time, the first number of customers waiting for service, and the first number of associates available.
12. The method of claim 11 , wherein the at least one machine learning process is based on XGBoost.
13. The method of claim 10 , further comprising:
determining at least one historic wait time corresponding to the arrival time window over a previous period of time;
determining a second number of customers waiting for service during the at least one historic wait time;
determining a second number of associates available during the at least one historic wait time; and
determining the expected wait time for the customer based on the at least one historic wait time, the second number of customers, and the second number of associates.
14. A non-transitory, computer-readable storage medium comprising executable instructions that, when executed by one or more processors, cause the one or more processors to:
receive first data identifying an arrival time window of a customer from a second computing device;
determine a current time based on the received first data;
determine a first number of customers waiting for service;
determine a first number of associates available to assist the first number of customers;
determine an estimated number of unexpected customers that may arrive during the arrival time window based on historical data, the historical data identifying numbers of unexpected customers during a same arrival time window for previous periods of time;
determine an expected wait time for the customer based on the current time, the first number of customers waiting for service, the estimated number of unexpected customers that may arrive during the arrival time window, and the first number of associates available; and
transmit the expected wait time to the second computing device.
15. The non-transitory computer-readable storage medium of claim 14 wherein the executable instructions, when executed by the one or more processors, cause the one or more processors to determine the expected wait time by applying at least one machine learning process to the current time, the first number of customers waiting for service, and the first number of associates available.
16. The non-transitory computer-readable storage medium of claim 14 wherein the executable instructions, when executed by the one or more processors, cause the one or more processors to:
determine at least one historic wait time corresponding to the arrival time window over a previous period of time;
determine a second number of customers waiting for service during the at least one historic wait time;
determine a second number of associates available during the at least one historic wait time; and
determine the expected wait time for the customer based on the at least one historic wait time, the second number of customers, and the second number of associates.
17. The non-transitory computer-readable storage medium of claim 14 wherein the executable instructions, when executed by the one or more processors, cause the one or more processors to determine the estimated number of unexpected customers by applying a machine learning process to the historical data identifying numbers of customers arriving unexpectedly during the previous periods of time.