Systems and methods for handling latent anomalies
Systems and methods for handling latent anomalies in field devices are described herein. When an anomaly is detected, the system can earmark the presence of the detected anomaly with a flag or other notification, and announce the existence of the anomaly to a user. In some embodiments, a self-test may be distributed to devices in the field that may be potentially affected by the latent anomaly so that those devices can monitor for the presence of the anomaly and take appropriate action if detected.
1. A method for determining existence of a latent anomaly, comprising:
performing a plurality of tests on a device using a test system, wherein each of the plurality of test has an expected result;
monitoring the device in accordance with a first one of the plurality of tests;
obtaining data during the first one of the plurality of tests; and
in response to determining a test result based on the obtained data is different than the expected result:
analyzing the data obtained during the first one of the plurality of tests to ascertain potential occurrence of a latent anomaly in the device.
2. The method of claim 1 , in response to detecting existence of the latent anomaly:
identifying a subset of a plurality of field devices that has potential to exhibit the latent anomaly; and
provide a self-test to each field device in the identified subset such that each field device in the identified subset can monitor for the latent anomaly and perform an action in response to monitoring for the occurrence of the latent anomaly.
3. The method of claim 1 , wherein the plurality of tests subject the device to stresses that field devices could potentially incur over their operational lives.
4. The method of claim 1 , wherein the obtaining data during the first one of the plurality of tests comprises:
obtaining an image of the device;
preparing the image for area of interest selection;
isolating at least one generic region of interest in the prepared image;
identifying an object in each of the at least one generic region of interest;
re-aligning the area of interest selection with the identified object; and
testing the device in accordance with the first one of the plurality of tests by observing the re-aligned area of interest selection.
5. The method of claim 4 , wherein the identifying comprises using shape matching to identify the object.
6. The method of claim 5 , wherein the using shape matching comprises:
applying a combination of Gaussian and Laplacian functions to produce an edge centric image; and
computing Hu's image moments on the edge centric image.
7. The method of claim 4 , wherein the identifying comprises using feature matching to identify the object.
8. The method of claim 7 , wherein the using feature matching comprises extracting features from the image using an accelerated-KAZE technique to identify the object.
9. The method of claim 1 , wherein the obtaining data during the first one of the plurality of tests comprises:
obtain a frame of video of the device;
extract pixels from at least one area of interest within the frame;
determine a state of the at least one area of interest based on the extracted pixels, wherein the state is used to determine the test result.
10. A method for handling latent anomalies, comprising:
evaluating a plurality of data sources to detect existence of a latent anomaly; and
in response to detecting existence of the latent anomaly:
in response to confirming that the latent anomaly does not prevent desired operation of a field device and cannot be corrected via an software update, providing a self-test to the field device such that the field device can monitor for the latent anomaly and perform an action in response to monitoring the occurrence of the latent anomaly.
11. The method of claim 10 , wherein in response to confirming that the latent anomaly is fatal to operation of the field device, providing a notice to a user of the field device that the latent anomaly fatal to field device operation.
12. The method of claim 10 , wherein in response to detecting existence of the latent anomaly can be corrected by a software update, instructing the field device to perform a software update.
13. The method of claim 10 , wherein the plurality of data sources comprises direct inputs that identify latent anomalies that are actually observed.
14. The method of claim 10 , wherein the plurality of data sources comprises
empirical inputs that identify latent anomalies that are empirically observed in data collected from field devices.
15. The method of claim 10 , wherein the plurality of data sources comprises a testing system capable of running a battery of tests that subject the field device to stresses that the field device could potentially incur over its operational life.
16. The method of claim 15 , wherein one of the test performed by the testing system comprises a visual observation test that monitors at least one area of interest on a test device.
17. The method of claim 15 , wherein one of the tests performed by the testing system comprises an audio observation test that verifies accuracy of speech emitted by the test device.
18. The method of claim 10 , wherein the latent anomaly is a defect that is discovered to exist within a field device after that field device has been sold.