node-red-contrib-elipsa-predictive-maintenance-cloud 1.0.1
Node-RED node for elipsa-predictive-maintenance-cloud
node-red-contrib-elipsa-predictive-maintenance-cloud
Node-RED node for elipsa-predictive-maintenance-cloud
Install
To install the stable version use the Menu - Manage palette - Install
option and search for node-red-contrib-elipsa-outlier-template-router, or run the following
command in your Node-RED user directory, typically ~/.node-red
npm install node-red-contrib-elipsa-predictive-maintenance-cloud
Usage
To utilize the Elipsa Predictive Maintenance node, you must first sign up for a free API Access Key @ www.elipsa.ai
Simple, Fast, and Flexible Predictive Maintenance
Parameters:
- Access Key: User's access key found on the profile page of the Elipsa platform
- API Key: The API Key of the Elipsa template that you will be streaming to
- Unique ID: Unique ID of the machine/device being monitored. (Ex: the serial number of the machine/device)
- Features: A comma separated list of the keys that contain your telemetry data in the msg.payload
Input:
payload
A Javascript Array of objects where each object contains the individual features and values of the current machine data that you are monitoring
Output:
payload
A Javascript object containing the following keys
prediction: Prediction as to whether the current data indicates abnormal operation of the machine/device being monitored. 0 indicates normal behavior with a response of 1 indicating an outlier
confidence: Confidence level in the prediction of an outlier. A confidence of 0 is returned when the prediction is 0
drivers: A list of values for each of the features associated with the template. Each feature has a value explaining that feature's influence in determining a predicted outlier. The drivers can be utilized to help diagnose the location of the issue
msg: OK if the API is running otherwise an error message is returned
status_code: Numerical status code indicating the current stage of the model lifecyle:
*0-Ready to Stream Data training *1-Streaming data before initial model build *2-Initial model training *3-Making predictions on the initial model and queuing for final model training *4-Final model *5-Making predictions on final model
status: Status message associated with status
Details:
Stream new telemetry data to a given Elipsa Rapid Deployment Template to monitor your machines and devices using AI-based Predictive Maintenance
The node utilizes the unique ID parameter to automate the AI model creation, deployment, and inference.
Simply connecting a machine to the node and future streaming data enables self-training no-touch outlier detection.
The input msg.payload must contain an array of objects for the node to work properly. The node takes the input msg.payload and extracts the values of the keys specified in features parameter
The node will send an http request to the Elipsa Platform sending the telemetry data and routing it appropriately based on the specified Access Key, API Key, and Unique ID
Elipsa Rapid Deployment Templates enable AI at scale. Monitor like machines through a single API endpoint. Elipsa will automate the creation and deployment of a predictive maintenace model specific to the defined unique ID
Example Response:
json
{
"prediction": "1.0",
"confidence": "0.63",
"drivers": {
"feature1_exp": 1.7,
"feature2_exp": 0.8,
"feature3_exp": 0.1,
},
"msg": "OK",
"status": "Streaming Data Against Initial Model. 6373 more data points required to retrain final model",
"status_code": 3
}
