Intelligent prediction of bundles of spare parts
Techniques for intelligently predicting bundles of replacement parts. These techniques include determining a plurality of maintenance events for a plurality of replacement parts. The determining includes identifying one or more replacement parts for a maintenance event, based on one or more replacement part events occurring within a time period related to the maintenance event. The techniques further include generating one or more clusters of replacement parts based on the plurality of maintenance events, and predicting one or more bundles of replacement parts, based on the clusters, wherein each bundle comprises a plurality of replacement parts.
1. A method for bundling parts, the method, comprising:
determining, using a computer processor, a plurality of maintenance events for a plurality of replacement parts, the determining comprising:
identifying, using the computer processor, one or more replacement parts for a maintenance event, based on one or more replacement part events occurring within a time period related to the maintenance event, wherein the time period relating to the maintenance event comprises a time period relating to a starting point in time for the maintenance event and an ending point in time for the maintenance event;
generating, using the computer processor, a two dimensional array relating to the plurality of maintenance events and the plurality of replacement parts;
generating, using the computer processor, one or more clusters of replacement parts based on the plurality of maintenance events and the two dimensional array;
calculating a local weight for a plurality of entries in the two dimensional array;
calculating a global weight for the plurality of entries in the two dimensional array;
reducing a dimension of the two dimensional array, based at least in part on the calculated local weights and global weights;
predicting, using the computer processor, one or more bundles of replacement parts, based on the clusters, wherein each bundle comprises a plurality of replacement parts necessary to complete a task;
bundling together, in each bundle as predicted, the plurality of replacement parts; and
responding to receiving a search comprising a replacement part in the bundle or of the task, by transmitting a bundle as predicted through an output device.
2. The method of claim 1 , wherein the one or more replacement part events comprise sales of replacement parts.
3. The method of claim 1 , wherein determining, using the computer processor, the plurality of maintenance events for the plurality of replacement parts further comprises generating time series data related to historical data relating to the plurality of replacement parts.
4. The method of claim 3 , further comprising calculating a kernel density estimate related to the time series data, wherein the time period related to the maintenance event is based on the kernel density estimate.
5. The method of claim 4 , wherein the two dimensional array is a binary array.
6. The method of claim 4 , wherein the one or more replacement part events comprise electronic searches for replacement parts.
7. The method of claim 1 , further comprising:
identifying, using the computer processor, a plurality of interchangeable parts among the plurality of replacement parts, wherein the plurality of maintenance events is based on the plurality of interchangeable parts.
8. The method of claim 7 , further comprising:
determining, using the computer processor, for a first interchangeable part of the plurality of interchangeable parts, an original part identifier and a most recent part identifier, wherein the plurality of maintenance events uses the most recent part identifier in place of the original part identifier.
9. The method of claim 1 , further comprising:
determining a price for a first bundle of the one or more bundles of replacement parts, based at least in part on pre-determined prices for each part in the first bundle, wherein the price for the first bundle is lower than combined pre-determined prices for each part in the first bundle.
10. The method of claim 9 , wherein the price for the first bundle is determined automatically, using the computer processor, and wherein the price for the first bundle is based on one or more characteristics of parts in the first bundle.
11. A system that comprises:
a processor; and
a program stored in a memory and configured to determine a plurality of maintenance events for a plurality of replacement parts based upon operations executed by the processor that comprise:
identifying one or more replacement parts for a maintenance event, based on one or more replacement part events occurring within a time period related to the maintenance event, wherein the time period relating to the maintenance event comprises a time period relating to a starting point in time for the maintenance event and an ending point in time for the maintenance event;
generating a two dimensional array relating to the plurality of maintenance events and the plurality of replacement parts;
generating one or more clusters of replacement parts based on the plurality of maintenance events and the two dimensional array;
calculating a local weight for a plurality of entries in the two dimensional array;
calculating a global weight for the plurality of entries in the two dimensional array;
reducing a dimension of the two dimensional array, based at least in part on the calculated local weights and global weights;
predicting one or more bundles of replacement parts, based on the clusters, wherein each bundle comprises a plurality of replacement parts necessary to complete a task;
bundling together, in each bundle as predicted, the plurality of replacement parts; and
responding to receiving a search, comprising a replacement part in the bundle or of the task, by transmitting the bundle as predicted through an output device.
12. The system of claim 11 , wherein the one or more replacement part events comprise at least one of sales of replacement parts or electronic searches for replacement parts.
13. The system of claim 11 , wherein determining the plurality of maintenance events for the plurality of replacement parts further comprises:
generating time series data related to historical data relating to the plurality of replacement parts; and
calculating a kernel density estimate related to the time series data, wherein the time period related to the maintenance event is based on the kernel density estimate.
14. The system of claim 11 , wherein the program is further configured to identify a plurality of interchangeable parts among the plurality of replacement parts, wherein the plurality of maintenance events is based on the plurality of interchangeable parts.
15. The system of claim 11 , wherein the program is further configured to determine a price for a first bundle of the one or more bundles of replacement parts, based at least in part on pre-determined prices for each part in the first bundle, wherein the price for the first bundle is lower than combined pre-determined prices for each part in the first bundle.
16. The system of claim 15 , wherein the price for the first bundle is determined automatically and wherein the price for the first bundle is based on one or more characteristics of parts in the first bundle.
17. A computer program product, that comprises a non-transitory computer-readable storage medium that comprises a computer-readable program code embodied therewith configured to execute operations on one or more computer processors, that determine a plurality of maintenance events for a plurality of replacement parts based upon operations that comprise:
identifying one or more replacement parts for a maintenance event, based on one or more replacement part events occurring within a time period related to the maintenance event, wherein the time period relating to the maintenance event comprises a time period relating to a starting point in time for the maintenance event and an ending point in time for the maintenance event;
generating a two dimensional array relating to the plurality of maintenance events and the plurality of replacement parts;
generating one or more clusters of replacement parts based on the plurality of maintenance events and the two dimensional array;
calculating a local weight for a plurality of entries in the two dimensional array;
calculating a global weight for the plurality of entries in the two dimensional array;
reducing a dimension of the two dimensional array, based at least in part on the calculated local weights and global weights;
predicting one or more bundles of replacement parts, based on the clusters, wherein each bundle comprises a plurality of replacement parts necessary to complete a task;
bundling together, in each bundle as predicted, the plurality of replacement parts; and
responding to receiving a search, comprising a replacement part in the bundle or of the task, by transmitting the bundle as predicted through an output device.
18. The computer program product of claim 17 , wherein the one or more replacement part events comprise at least one of sales of replacement parts or electronic searches for replacement parts.
19. The computer program product of claim 17 , the operations further comprising:
determining a price for a first bundle of the one or more bundles of replacement parts, based at least in part on pre-determined prices for each part in the first bundle, wherein the price for the first bundle is lower than combined pre-determined prices for each part in the first bundle.
20. The computer program product of claim 19 , wherein the price for the first bundle is determined automatically and wherein the price for the first bundle is based on one or more characteristics of parts in the first bundle.