Scalable systems of site inventory scanners and vehicles for dynamic assessment, deployment and reconciliation of inventories

1 . A dynamic assessment, deployment and reconciliation system of a plurality of different, non-durable stock-keeping units (SKUs) for a plurality of different retail facilities in a plurality of different geographic locations, the system comprising:

a processor;

a network interface, configured for data communication over a network, coupled to the processor;

a memory coupled to the processor;

programming in the memory, wherein execution of the programming by the processor configures the system to implement functions, including, for each of a plurality of different target SKUs for different, non-durable SKUs, functions to periodically:

receive a target SKU count input of the target SKU;

over the network interface, receive an actual SKU count input of the target SKU on an assessment date;

determine a future delivery date, the future delivery date later than the assessment date;

calculate a date difference, the date difference based on a difference between the assessment date and the future delivery date;

calculate a count difference, the count difference based on a difference between the target SKU count input and the actual SKU count input;

simulate, using a trained model, an effect of time on a quantity of target SKU required for delivery, using the date difference and the count difference, to determine a projected target SKU count;

based on the projected target SKU count, issue a packing instruction, the packing instruction configured to instruct a packing of the target SKU onto a transportation vehicle at a dynamically selected intermediate facility;

based on an actual packing outcome of the transportation vehicle at the dynamically selected intermediate facility, issue a first stop delivery instruction, the first stop delivery instruction including the first retail facility and a first delivery amount of the target SKU;

over the network interface, receive a first stop delivery outcome, the first stop delivery outcome reporting a first delivered amount of the target SKU differing from the first delivery amount as a first delivery difference; and

simulate, using the trained model, an effect of the first delivery difference on a quantity of the target SKU required for delivery to a second retail facility of the plurality of different retail facilities in the same geographical area, to dynamically determine an updated quantity of the target SKU required for delivery to the second retail facility.

2 . The system of claim 1 , wherein:

the trained model is periodically trained based on a training data set; and

the training data set includes at least one data point wherein a data point date difference of the data point is greater or lesser than the date difference.

3 . The system of claim 1 , wherein execution of the programming by the processor further configures the system to implement functions, including, for each of the plurality of different target SKUs for different, non-durable SKUs, functions to periodically:

receive a stocking results report over the network interface, the stocking results report describing a difference between the projected target SKU count and an actual SKU quantity delivered; and

re-train the trained model based on the stocking results report.

4 . The system of claim 1 , wherein the functions are further periodically implemented for each of a plurality of different retail facilities in each of a plurality of different geographic locations.

5 . The system of claim 1 , wherein:

execution of the programming by the processor further configures the system to implement functions, including, for each of the plurality of different target SKUs for different, non-durable SKUs, functions to periodically:

over the network interface, for each of a plurality of different retail facilities in the same geographical area, receive a respective actual SKU count input of the target SKU on a respective assessment date;

wherein the count difference is further based on a difference between the target SKU count input and a sum of the respective actual SKU count inputs of the target SKU; and

the packing instruction further includes respective delivery quotas associated with each of the respective actual SKU count inputs of the target SKU.

6 . The system of claim 1 , wherein

determining the first delivery amount of the target SKU and the updated quantity of the target SKU required for delivery to the second retail facility further comprises:

simulating, using the trained model, a plurality of candidate allocations of the target SKU among the first retail facility and the second retail facility, each candidate allocation associated with at least one performance metric comprising at least one of: a predicted under-delivery probability, a predicted over-delivery probability, or an expected transportation cost; and

selecting, as the first delivery amount and the updated quantity, the allocation that optimizes the at least one performance metric subject to one or more delivery constraints.

7 . The system of claim 1 , wherein the first delivery amount of the target SKU is the projected target SKU count, and the actual SKU count input is captured at the first retail facility of the plurality of different retail facilities in the same geographical area on the assessment date.

8 . The system of claim 1 , further comprising a scanner device, wherein the actual SKU count input of the target SKU on the assessment date is configured to be received from the scanner device operated at an on-site location of the first retail facility.

9 . The system of claim 1 , wherein:

the first delivery amount of the target SKU, and the updated quantity of the target SKU required for delivery to the second retail facility of the plurality of different retail facilities in the same geographical area, are dynamically determined before the packing instruction is issued.

10 . The system of claim 1 , wherein

the assessment date is on an assessment schedule; and

execution of the programming by the processor further configures the system to implement functions, including functions to periodically:

simulate, using the trained model, an effect of the first delivery difference on the assessment schedule, to determine an updated assessment schedule.

11 . The system of claim 1 , wherein:

execution of the programming by the processor further configures the system to implement functions, including functions to periodically:

issue the first stop delivery instruction, the first stop delivery instruction including the first retail facility of a plurality of different retail facilities in the same geographical area, and the first delivery amount of the target SKU, and

issue a second stop delivery instruction, the second stop delivery instruction including the second retail facility of the plurality of different retail facilities in the same geographical area, and the updated quantity of the target SKU as a second delivery amount of the target SKU; and

the projected target SKU count includes the first delivery amount and the second delivery amount.

12 . The system of claim 11 , wherein:

the first stop delivery instruction further includes a first stop target time;

the second stop delivery instruction further includes a second stop target time; and

execution of the programming by the processor further configures the system to implement functions, including functions to periodically:

simulate, using the trained model, an effect of the first stop delivery instruction and the second stop delivery instruction on one or more transportation vehicles at one or more intermediate facilities servicing the plurality of different retail facilities in the same geographical area, to determine a delivery route.

13 . A method of assessing, deploying, and reconciling a plurality of different, non-durable stock-keeping units (SKUs) for a plurality of different retail facilities in a plurality of different geographic areas, the method comprising:

receiving a target SKU count input of a target SKU;

receiving an actual SKU count input of the target SKU on an assessment date;

determining a future delivery date, the future delivery date later than the assessment date;

calculating a date difference, the date difference based on a difference between the assessment date and the future delivery date;

calculating a count difference, the count difference based on a difference between the target SKU count input and the actual SKU count input;

simulating, using a trained model implementing sales velocity based on multi-period historical consumption rates, an effect of time on a quantity of target SKU required for delivery, using the date difference and the count difference, to determine a projected target SKU count, the projected target SKU count representing a predicted change in stock of the target SKU at a retail facility between the assessment date and the future delivery date; and

based on the projected target SKU count, issuing a packing instruction, the packing instruction configured to instruct a packing of the target SKU onto a transportation vehicle at a dynamically selected intermediate facility.

14 . The method of claim 13 , further comprising:

receiving a stocking results report, the stocking results report describing a difference between the projected target SKU count and an actual SKU quantity delivered; and

re-training the trained model based on the stocking results report.

15 . The method of claim 13 , further comprising:

for each of a plurality of different retail facilities in the same geographical area, receiving a respective actual SKU count input of the target SKU on a respective assessment date;

wherein:

the count difference is further based on a difference between the target SKU count and a sum of the respective actual SKU count inputs of the target SKU; and

the packing instruction further includes respective delivery quotas associated with each of the respective actual SKU count inputs of the target SKU.

16 . The method of claim 13 , further comprising:

based on an actual packing outcome of the transportation vehicle at the dynamically selected intermediate facility, issuing a first stop delivery instruction, the first stop delivery instruction including a first retail facility of a plurality of different retail facilities in the same geographical area and a first delivery amount of the target SKU.

17 . The method of claim 16 , wherein the first delivery amount of the target SKU is the projected target SKU count, and the actual SKU count input is captured at the first retail facility of the plurality of different retail facilities in the same geographical area on the assessment date.

18 . The method of claim 16 , further comprising:

receiving a first stop delivery outcome, the first stop delivery outcome reporting a first delivered amount of the target SKU differing from the first delivery amount as a first delivery difference;

simulating, using the trained model, an effect of the first delivery difference on a quantity of the target SKU required for delivery to a second retail facility of the plurality of different retail facilities in the same geographical area, to dynamically determine an updated quantity of the target SKU required for delivery to the second retail facility; and

simulating, using the trained model, an effect of the first delivery difference on an assessment schedule, to determine an updated assessment schedule;

wherein:

the first delivery amount of the target SKU, and the quantity of the target SKU required for delivery to the second retail facility of the plurality of different retail facilities in the same geographical area, are dynamically determined before the packing instruction is issued; and

the assessment date is on the assessment schedule.

19 . The method of claim 13 , further comprising:

issuing a first stop delivery instruction, the first stop delivery instruction including a first retail facility of a plurality of different retail facilities in the same geographical area, and a first delivery amount of the target SKU, and

issuing a second stop delivery instruction, the second stop delivery instruction including a second retail facility of the plurality of different retail facilities in the same geographical area, and a second delivery amount of the target SKU;

wherein the projected target SKU count includes the first delivery amount and the second delivery amount.

20 . The method of claim 19 , further comprising:

simulating, using the trained model, an effect of the first stop delivery instruction and the second stop delivery instruction on one or more transportation vehicles at one or more intermediate facilities servicing the plurality of different retail facilities in the same geographical area, to determine a delivery route;

wherein:

the first stop delivery instruction further includes a first stop target time; and

the second stop delivery instruction further includes a second stop target time.