Building analysis system with machine learning based interpretations

1. A vibration analysis system for predicting performance of building equipment, the system comprising:

one or more memory devices configured to store instructions that, when executed on one or more processors, cause the one or more processors to:

receive vibration data from the building equipment, the vibration data indicating the performance of the building equipment;

generate, based on a prediction model and the vibration data, a performance prediction for the building equipment;

generate, based on the vibration data, a performance explanation for the performance prediction, the performance explanation comprising an indication of one or more data points of the vibration data associated with the performance prediction and further comprising an explanation indicating one or more reasons that the performance prediction will occur; and

cause a user interface to display the performance prediction and the performance explanation.

2. The vibration analysis system of claim 1 , wherein the instructions cause the one or more processors to:

receive prior performance data of the building equipment; and

train the prediction model based on the received prior performance data of the building equipment.

3. The vibration analysis system of claim 1 , wherein the instructions cause the one or more processors to:

receive prior performance data of other building equipment; and

train the prediction model based on the received prior performance data of the other building equipment.

4. The vibration analysis system of claim 1 , wherein the instructions cause the one or more processors to:

receive, via the user interface, user feedback from a user, the user feedback based on the performance prediction and the performance explanation; and

train the prediction model based on the user feedback.

5. The vibration analysis system of claim 4 , wherein the instructions cause the one or more processors to:

receive a second vibration data from the building equipment;

generate, based on the trained prediction model, a second performance prediction for the building equipment, the second performance prediction taking into account the user feedback;

generate, based on the second vibration data, a second performance explanation for the second performance prediction, the second performance explanation comprising an indication of one or more data points of the second vibration data associated with the second performance prediction and further comprising an explanation indicating one or more reasons that the second performance prediction will occur, the second performance prediction based on the user feedback; and

cause a user interface to display the second performance prediction and the second performance explanation.

6. The vibration analysis system of claim 4 , wherein the user interface comprises selection boxes that the user can interact with to indicate agreement or disagreement with at least one of the performance prediction and the performance explanation.

7. The vibration analysis system of claim 4 , wherein the user interface comprises a text box in which the user can enter text to indicate reasons for agreement or disagreement with at least one of the performance prediction and the performance explanation.

8. The vibration analysis system of claim 4 , wherein the user interface comprises a chart indicating the vibration data on which at least one of the performance prediction and the performance explanation is based.

9. The vibration analysis system of claim 8 , wherein the chart comprises a highlighted area, the highlighted area indicating the one or more data points, wherein values of the one or more data points cause the performance prediction.

10. The vibration analysis system of claim 1 , wherein the user interface displays the performance explanation as text.

11. The vibration analysis system of claim 1 , wherein the instructions cause the one or more processors to generate an audio message indicating the performance explanation.

12. The vibration analysis system of claim 1 , the performance prediction comprising a prediction of the performance of a component of the building equipment.

13. A method for predicting performance of building equipment, the method comprising:

receiving, by a processing circuit, vibration data from the building equipment, the vibration data indicating the performance of the building equipment;

analyzing, by the processing circuit, the vibration data with a data analysis model;

generating, by the processing circuit, a performance prediction for the building equipment, the performance prediction based on the analyzed vibration data and a prediction model;

generating, by the processing circuit, a performance explanation for the performance prediction, the performance explanation comprising an indication of one or more data points of the vibration data associated with the performance prediction and further comprising an explanation indicating one or more reasons that the performance prediction will occur; and

displaying, on a user interface by the processing circuit, the performance prediction and the performance explanation to a user.

14. The method of claim 13 , wherein the receiving step further comprises:

receiving, by the processing circuit, prior performance data of the building equipment, and;

training, by the processing circuit, the prediction model based on the received prior performance data of the building equipment.

15. The method of claim 13 , wherein the receiving step further comprises:

receiving, by the processing circuit, prior performance data of other building equipment; and

training, by the processing circuit, the prediction model based on the received prior performance data of the other building equipment.

16. The method of claim 13 , further comprising:

receiving, by the processing circuit via the user interface, user feedback from a user, the user feedback based on the performance prediction and the performance explanation; and

training, by the processing circuit, the prediction model based on the user feedback.

17. The method of claim 16 , further comprising:

receiving, by the processing circuit, a second vibration data from the building equipment;

generating, by the processing circuit, based on the trained prediction model, a second performance prediction for the building equipment, the second performance prediction based on the user feedback;

generating, by the processing circuit, based on the second vibration data, a second performance explanation for the second performance prediction, the second performance explanation comprising an indication of one or more data points of the second vibration data associated with the second performance prediction and further comprising an explanation indicating one or more reasons that the second performance prediction will occur, the second performance prediction taking into account the user feedback; and

displaying, on the user interface by the processing circuit, the second performance prediction and the second performance explanation.

18. The method of claim 17 , wherein the user interface comprises selection boxes that the user can interact with to indicate agreement or disagreement with the performance prediction and/or the performance explanation.

19. The method of claim 17 , wherein the user interface comprises a text box in which the user can enter text to indicate reasons for agreement or disagreement with at least one of the performance prediction and the performance explanation.

20. A building system of a building, the system comprising:

building equipment configured to operate to control one or more environmental conditions of the building, wherein the building equipment comprise one or more vibration sensors configured to collect vibration data while the building equipment operates to control the one or more environmental conditions;

a processing circuit configured to:

receive the vibration data from the building equipment, the vibration data indicating performance of the building equipment;

generate, based on a prediction model and the vibration data, a performance prediction for the building equipment;

generate, based on the vibration data, a performance explanation for the performance prediction, the performance explanation comprising an indication of one or more data points of the vibration data associated with the performance prediction and further comprising an explanation indicating one or more reasons that the performance prediction will occur; and

cause a user interface to display the performance prediction and the performance explanation.