The Advanced Fault Detection (AFD) API is used to detect anomalies in complex and repetitive systems. For instance, it could be used to anticipate failures in pump motors, drives or robotic arms. It leverages unsupervised machine learning techniques to model the expected normal behavior of a system. Then, it uses this model to detect abnormal behavior that allows to anticipate a failure of the system. This document provides a general tutorial for users who want to consume the Advanced Fault Detection API.

The Advanced Fault Detection API includes 4 services:

• Learn the model from historical data (createModel).
• Calibrate the created model to the desired false positive rate (finalizeModel).
• Apply the model on new data and detect faults (applyModel).
• Get information on an existing model (getModelInformation)

Take the tour and experiment on your own data prior to any development through our hands-on application! https://try-analytics-se.azurewebsites.net/afd

Before you can sign up for the Advanced Fault Detection API, you must register or login with an Exchange account. Then you can subscribe to the API product in our Shop and where you can read about terms & conditions. After subscription, you will find your API key prepopulated in the Try Out console.

1. Select one of the endpoints and click on
2. Select the Production as the target environment
3. Your API key is shown in the Try Out console Status of your API consumption will soon be available from the Exchange cockpit

The createModel endpoint learns a fault-detection model choosing the best parameters for the selected anomaly-detection algorithm applied over a training data. This service takes as training data just normal observations, i.e. it is assumed that no anomalies are present.

The finalizeModel endpoint takes an existing model from createModel service and finalizes it by providing a threshold for a false positive ratio provided by the user.

The applyModel endpoint takes a model previously trained and finalized (it is compulsory for the model to have passed through finalizeModel) and applies it over a new dataset having the same schema. The service returns the input dataset plus three additional columns containing indicators functions, two of them signal a faulty observation with a value equal to one, and the other one provides the score of each observation; additionally this service provides the threshold for anomaly detection provided previously by finalizeModel endpoint.

The getModelInformation endpoint returns information on an existing model that is securely stored on the Schneider Electric platform.

Amount of calls to the API in the SANDBOX are limited. To get a full experience and extend the thresholds, please subscribe to the API and use the production environment.

Step 1: Setup environment

Step 2: Prepare Data

Step 3: Create a new model

Step 4: Finalize model

Step 5: Apply an existing Advanced Fault Detection model to new data

Step 6: Get model information on an existing Advanced Fault Detection model

This API uses API keys to authenticate and allow access to the API. Your API keys carry many privileges, so be sure to keep them secure! The API key should be included in all API requests in the Authorization header that looks like the following: Authorization: your_subscription_key Please refer the 'How to sign up for the API' section under Developer Guide above to know how to get your subscription key.

We follow the error response format proposed in RFC 7807also known as Problem Details for HTTP APIs. As with our normal API responses, your client must be prepared to gracefully handle additional members of the response.

Contact the SE Exchange support team at exchange.support@se.com and include:

• Endpoint URL
• Request/Response
• Any additional information like Screenshots , Postman collections.

Authentication is the act of proving an assertion, such as the identify of a computer system user.