Amazon Technologies, Inc.

Llewellyn W Bezanson

Girish S Lakshman

Saurabh Dileep Baji

Pradeep Desai

Benjamin Elliott Pew

M Christopher Wenneman

Keith B Ligon

A method and system for transit characteristic prediction. In one embodiment, a method may include determining respective transit latencies from a source location to a number of destination locations, and grouping the destination locations according to a similarity criterion into a number of subsets corresponding to respective geographical regions, where transit characteristics of locations grouped into a geographical region may each satisfy the similarity criterion. The method may further include determining a respective distribution associated with each of the geographical regions, where a given respective distribution correlates a given transit characteristic from the source location to a given geographical region with a respective surety factor, where the respective surety factor includes a cumulative probability that the given transit characteristic will be achieved. The method may also include dynamically updating the respective transit characteristic, the grouping of destination locations into geographical regions and the respective distributions dependent upon collecting empirical transit data.

Method and system for transit characteristic prediction

A method and system for transit characteristic prediction. In one embodiment, a method may include determining respective transit latencies from a source location to a number of destination locations, and grouping the destination locations according to a fitness function into a number of subsets corresponding to respective geographical regions. The grouping may involve a series of divisions and combinations of potential regions to form a plurality of sets of potential regions. Each set of potential regions may be evaluated using the fitness function, and the set with the better fitness score may be selected. The method may also include dynamically updating the respective transit characteristic, regrouping the regions, and reselecting a set of potential regions based on empirical transit data.

Adaptive regionalization for transit characteristic prediction

A method and system for selecting a shipping method dependent upon a dynamic model of shipping activity. In one embodiment, a method may include detecting a given shipment event for which one of a plurality of shipping methods is to be selected, determining a current output state of a discrete event model representative of shipping method selections for one or more previous shipment events, and in response to detecting the given shipping event, producing a selection of a particular one of the plurality of shipping methods that mathematically optimizes an objective function, wherein the objective function is dependent on a cost of shipping associated with the given shipment event and on a weighted deviation between the current output state of the discrete event model and a target output state of the discrete event model.

Growing community of inventors

North Bend, WA, United States of America

Llewellyn W Bezanson