System and method for estimating delivery events

• US 9,508,053 B2
• Filed: 10/18/2013
• Issued: 11/29/2016
• Est. Priority Date: 10/18/2013
• Status: Active Grant

• Current Assignee: Caterpillar Incorporated
• Original Assignee: Caterpillar Incorporated
• Inventors: Collins, Darryl Victor
• Primary Examiner(s): Rennema, Robert
• Assistant Examiner(s): Mia, Md N
• Application Number:
  US14/057,462
• Publication Number:
  US 20150112954A1
• Time in Patent Office: 1,138 Days
• Field of Search: 701/50, 701/124, 701/472, 701/53, 701/54, 701/84, 701/87, 701/90
• US Class Current: 1/1
• CPC Class Codes: G06Q 10/083 Shipping

Abstract

A computer-implemented method for processing loading operation data to estimate one or more delivery events is described. The loading operation data comprises payload records, each representing a delivery event. The method comprises, for a selected payload record, operating a processing unit to determine whether a delivery weight in respect of the selected payload record is likely to reflect a single delivery event. If the delivery weight is unlikely to reflect a single delivery event the method further comprises generating a synthetic payload record representing an estimated delivery event and comprising synthetic weight and timing values.

Claims

1. A method for processing loading operation data to estimate one or more delivery events, the loading operation data comprising a plurality of payload records, each payload record representing a delivery event, wherein the method comprises:

delivering material from a loading machine into a haulage machine thereby generating payload messages; and

for a selected payload record, operating a processing unit to;

determine whether a delivery weight in respect of the selected payload record is likely to reflect a single delivery event, the determination being based at least in part on an estimated loading machine payload weight;

if the delivery weight is unlikely to reflect a single delivery event, generate a synthetic payload record representing an estimated delivery event, the synthetic payload record comprising;

a weight value representing an estimated haulage machine payload weight at the estimated delivery event, and

a timing value representing an estimated time at which the weight value for the synthetic payload record would have occurred, and

update the loading operation data by incorporating the synthetic payload record into the loading operation data;

wherein the updated loading operation data provides a more accurate record of loading operations for which the payload messages may not accurately report.

2. A method according to claim 1, wherein determining whether the delivery weight in respect of the selected payload record is likely to reflect a single delivery event comprises:

comparing the delivery weight with an acceptable value; and

if the delivery weight is greater than the acceptable value determining that the delivery weight is not likely to reflect a single delivery event.

3. A method according to claim 2, wherein the acceptable value is calculated so that the delivery weight in respect of the selected payload record will not be determined to be likely to reflect a single delivery event if the delivery weight exceeds the estimated loading machine payload weight by a statistically significant amount.

4. A method according to claim 1, wherein the estimated loading machine payload weight is calculated with reference to a loading machine payload weight dataset.

5. A method according to claim 4, wherein the estimated loading machine payload weight is one of the mode or the mean of values in the loading machine payload weight dataset.

6. A method according to claim 4, wherein the loading machine payload weight dataset is populated with delivery weights in respect of payload records.

7. A method according to claim 6, wherein the delivery weight of a given payload record is calculated by subtracting a weight value of an immediately preceding payload record in the loading operation data from the weight value of the given payload record.

8. A method according to claim 1, wherein the timing value is determined by reference to a relationship between the plurality of payload records.

9. A method according to claim 8, wherein the relationship is determined by performing a regression analysis of the plurality of payload records.

10. A method according to claim 1, further comprising operating the processing unit to re-implement the method of claim 1 using the loading operation data as updated with the synthetic payload record.

11. A method according to claim 1, wherein if the delivery weight is determined to be likely to reflect a single delivery event, and the selected payload record is an earliest payload record in the loading operation data, the method further comprises:

determining a difference between the weight value of the selected payload record and the estimated loading machine payload weight to be carry-back.

12. A method according to claim 1, wherein generating the synthetic payload record further comprises:

determining a weight uncertainty representing an uncertainty with respect to the weight value of the synthetic payload record; and

determining a timing uncertainty representing an uncertainty with respect to the estimated timing value of the synthetic payload record.

13. A method according to claim 12, wherein the weight uncertainty for the synthetic payload record is determined with reference to a measure of dispersion of values in a loading machine payload weight dataset.

14. A method according to claim 12, wherein the timing uncertainty for the synthetic payload record is determined with reference to the weight uncertainty and a relationship between the plurality of payload records.

15. A system comprising:

A loading machine delivering material from the loading machine into a haulage machine thereby generating payload messages; and

a non-transitory computer-readable medium comprising instructions which, when implemented by a computer processing system, cause the computer processing system to process loading operation data comprising a plurality of payload records, each payload record representing a delivery event, to estimate one or more delivery events by;

determining whether a delivery weight in respect of a selected payload record is likely to reflect a single delivery event, the determination being based at least in part on an estimated loading machine payload weight;

if the delivery weight is unlikely to reflect a single delivery event, generating a synthetic payload record representing an estimated delivery event, the synthetic payload record comprising;

a weight value representing an estimated haulage machine payload weight at the estimated delivery event, and

a timing value representing an estimated time at which the weight value for the synthetic payload record would have occurred, and

updating the loading operation data by incorporating the synthetic payload record into the loading operation data;

wherein the updated loading operation data provides a more accurate record of loading operations for which the payload messages may not accurately report.

16. A processing system comprising:

means for delivering material from a loading machine into a haulage machine thereby generating payload messages;

means for accessing loading operation data comprising a plurality of payload records, each payload record representing a delivery event;

means for selecting a payload record from the loading operation data;

means for determining whether a delivery weight in respect of the selected payload record is likely to reflect a single delivery event, the determination being based at least in part on an estimated loading machine payload weight;

means for, if the delivery weight is unlikely to reflect a single delivery event, generating a synthetic payload record representing an estimated delivery event, the synthetic payload record comprising;

a weight value representing an estimated haulage machine payload weight at the estimated delivery event, and

a timing value representing an estimated time at which the weight value for the synthetic payload record would have occurred, and

means for updating the loading operation data by incorporating the synthetic payload record into the loading operation data,

wherein the updated loading operation data provides a more accurate record of loading operations for which the payload messages may not accurately report.

17. A method for processing loading operation data to estimate one or more delivery events, the loading operation data comprising a plurality of payload records, each payload record representing a delivery event and comprising:

a weight value representing a weight of a haulage machine payload at a particular time; and

a timing value representing the particular time,

wherein the method comprises;

delivering material from a loading machine into a haulage machine thereby generating payload messages; and

operating a computer processing unit to;

compare a weight value of a selected payload record to an acceptable value, the acceptable value being based at least in part on an estimated loading machine payload weight;

if the weight value of the earliest existing payload record is greater than the acceptable value, generating a synthetic payload record representing an estimated earlier delivery event, the synthetic payload record comprising;

a weight value equal to a difference between the weight value of the selected payload record and the estimated loading machine payload weight, and

a timing value representing an estimated time at which the weight value for the synthetic payload record would have occurred, the timing value being based on a relationship between the plurality of payload records, and

updating the payload delivery data by incorporating the synthetic payload record into the payload data;

wherein the updated loading operation data provides a more accurate record of loading operations for which the payload messages may not accurately report.