Power your applications with Elipsa's library of no-touch AI and advanced analytics APIs. Our API-based approach provides the easiest way to build and deploy AI and advanced analytics at scale.
\nSign-up for a free access key: https://www.elipsa.ai/signup
\nFor questions and issues please contact us
\nOur APIs are broken into the following categories:
\nCreate and manage deployment templates to rapidly enable intelligent monitoring of your machines and devices.
\nEndpoint to return a list of all deployment templates that are associated with a specific access key.
\nList of templates with the following information:
\n[ \n {\n \"Name\": \"Vibration Analysis (Predictive Maintenance)\", \n \"Description\": \"Public Predictive Maintenance Rapid Deployment\n Template \",\n \"Features\": \"X,Y,Z\",\n \"ApiKey\": ,\n \"ModelPrefix\": \"Vibration-PM-\",\n \"InitialTrainAmt\": 2880, \n \"RetrainAmt\": 43200\n }\n]\n\nEndpoint to return a list of all models associated with a particular deployment template.
\nEndpoint returns a list of model objects in JSON formar with the following attributes:
\nSAMPLE RESPONSE:\n[\n {\n \"Name\": \"PM-VAV-XYZ123\",\n \"Features\": \"temp_room, temp_d, humidity, co2\", \n \"Status\": \"Streaming Data Against Initial Model. 39827 more data points required to retrain\n final model\"\n }, \n {\n \"Name\": \"PM-VAV-XYZ456\", \n \"Features\": \"temp_room, temp_d, humidity, co2\",\n \"Status\": \"Model Created: Ready To Stream Data\"\n }, \n ...\n]\n\nEndpoint to create a new Deployment Template associated with your user.
\nSandbox customers are limited to one saved template.
\n{\"msg\": \"MODEL CREATED: XYZ-123456-ABC>\"}\n\nEndpoint to stream new telemetry data to a given template (specifying the template's API Key in the url).
\nThe router utilizes the unique ID parameter to automate the AI model creation, deployment, and inference.
\nSimply connecting a machine to the template for future data streaming enables self-training no-touch outlier detection.
\n {\n \"prediction\": \"1.0\",\n \"confidence\": \"0.63\",\n \"drivers\": {\n \"feature1_exp\": 1.7,\n \"feature2_exp\": 0.8,\n \"feature3_exp\": 0.1,\n },\n \"msg\": \"OK\",\n \"status\": \"Streaming Data Against Initial Model. 6373 more data points required to retrain final model\",\n \"status_code\": 3\n}\n\nSelf-training deployment templates require no historical data to get started. Utilize our APIs to create a template based on the data that you are monitoring. Once created, you can stream new data to this template through our APIs and Elipsa will automate the process from there. The platform will aggregate the new data, build an initial model for each device being monitored, deliver predictions, and retrain models over time.
\n","_postman_id":"490cf143-52e3-4f3a-ba83-667ed7410ad6"},{"name":"Individual Models","item":[{"name":"Train and Save Outlier Detection Model","id":"b0e88945-3c83-44fa-befa-4cc8b3e57b1c","protocolProfileBehavior":{"disableBodyPruning":true,"disabledSystemHeaders":{"content-type":true,"content-length":true,"host":true,"accept":true,"accept-encoding":true,"connection":true}},"request":{"method":"POST","header":[{"key":"Content-Type","value":"application/json","type":"default"}],"body":{"mode":"raw","raw":"{\r\n \"Access_Key\":\"Train an outlier detection model on historical data to save for future predictions on new streaming data.
\nSaved models must be unique from other models that a user has saved. If the model name is not unique, or if an error occurs when building the model, the endpoint will return an error message in the 'msg' key. Successful model builds will return the following:
\n{\"msg\": \"Model Saved: Test-0\"}\n\nEndpoint to stream new telemetry data to a given saved model (specifying the model's API Key in the url) to predict whether the streaming data is an outlier.
\n{\n \"Prediction\": 1, \n \"Anomaly Score\": 0.9849821964003821, \n \"Drivers\": {\n \"amplitude_x_exp\": 1.3424207570871112,\n \"amplitude_y_exp\": 1.5740541054997308, \n \"amplitude_z_exp\": 2.602248723229603, \n \"amps_exp\": 1.5997539857099485\n }, \n \"API_Key\":\n\"828036f2-635a-11ec-8e7f-0acec021ad7c\"\n}\n\nEndpoint to analyze past data for outliers without saving a model for future predictions. Input historical data and determine which elements in the dataset constitute outliers.
\nThe response is a list of JSON objects consisiting of the input Data as well as the prediction as to whether the record is an outlier (0 for no and 1 for yes) and the anomaly score representing the confidence of the prediction for each record.
\n[{\"Bearing 1\": 0.0602364373260406, \"Bearing 2\": 0.0742266223936712, \"Bearing 3\": 0.0839260706089162, \"Bearing 4\":\n0.0444430880414066, \"Prediction\": 0, \"Anomaly Score\": 0.0}, {\"Bearing 1\": 0.0614554421602611, \"Bearing 2\":\n0.0738440841837977, \"Bearing 3\": 0.0844574442111431, \"Bearing 4\": 0.0450814004590059, \"Prediction\": 0, \"Anomaly Score\":\n0.0}, {\"Bearing 1\": 0.0613607598027244, \"Bearing 2\": 0.0756087699594707, \"Bearing 3\": 0.0828368084379122, \"Bearing 4\":\n0.0451183651545475, \"Prediction\": 1, \"Anomaly Score\": 0.2}, {\"Bearing 1\": 0.0616648273841493, \"Bearing 2\":\n0.0732786268860776, \"Bearing 3\": 0.0848792421504951, \"Bearing 4\": 0.0441722252063074, \"Prediction\": 1, \"Anomaly Score\":\n0.86}\n\nFor users who have collected historical data from their machines or devices, utilize the Elipsa APIs to quickly and easily build predictive models off of the historical data.
\nTrain and save models to make predictions against in the future as new data is collected or simply analyze existing datasets to instantly identify known outliers.
\n","_postman_id":"068119db-e3c1-4f10-affe-76ba23ea2dff"}],"id":"46305115-6757-40e6-8072-5f12a1b459eb","description":"Elipsa enables simple, fast, and flexible predictive maintenance using machine learning based outlier detection.
\nOur API driven approach enables no-touch creation, and deployment of predictive solutions that easily plug into your existing workflows.
\nWhether you have historical data or not, our various endpoints enable users to easily start monitoring their devices and data for outliers.
\n","_postman_id":"46305115-6757-40e6-8072-5f12a1b459eb"},{"name":"ADVANCED ANALYTICS","item":[{"name":"Geospatial Analysis","item":[{"name":"Asset Movement Insights","id":"063ea6e2-c28b-4ace-a1f5-04c201b92fd6","protocolProfileBehavior":{"disableBodyPruning":true,"disabledSystemHeaders":{"content-type":true,"content-length":true,"host":true,"accept":true,"accept-encoding":true,"connection":true}},"request":{"method":"POST","header":[{"key":"Content-Type","value":"application/json","type":"default"}],"body":{"mode":"raw","raw":"{\r\n \"Access_Key\":\"Endpoint to convert a collection of raw GPS coordinates into aggregated insights related to an asset's movement. The endpoint breaks the raw GPS data into movement and idle periods and provides analysis on the results.
\nAccess_Key: User's access key found on the profile page of the platform
\nGPS_String_Flag: Flag indicating whether the list of coordinates will be sent as a comma separated string (True) or as individual latitude and longitude values (False)
\nmin_idle_time: The number of seconds that an asset needs to sit idle before it is considered to have a new track started
\nData: list of GPS coordinate objects. Each coordinate object consists of the Date and Latitude/Longitude coordinates. The Date consists of date and time down to the closest second. The format of the coordinates differs based on the value of the GPS_String_Flag
\nAggregated insights/statistics from analyzing the provided coordinates:
\n{\n \"num_of_times_moved\": \"3\", \n \"total_move_distance\": \"65.61113456534744\", \n \"total_move_time\": \"31\", \n \"avg_move_dist\": \"21.87037818844915\", \n \"avg_move_time\": \"10.333333333333334\", \n \"max_move_dist\": \"33.574318610290405\", \n \"max_move_time\": \"17\", \n \"min_move_dist\": \"1.2008489725416527\", \n \"min_move_time\": \"1\", \n \"num_of_times_idle\": \"3\", \n \"total_idle_time\": \"28\",\n \"avg_idle_time\": \"9.333333333333334\", \n \"max_idle_time\": \"10\", \n \"min_idle_time\": \"9\", \n \"avg_speed\": \"2.116488211785401\",\n \"max_speed_per_hr\": \"0.5595719768381734\"\n}\n\nEndpoint to take in raw GPS coordinates and calculate insights around individual trips/tracks.
\nThe endpoint calculates individual tracks, and stops, within the list of GPS coordinates and returns a summary of statistics for each track and idle period.
\nList of aggregated tracks and detected stops from the GPS coordinates.
\nA track consists of the path of a moving asset until it comes to rest. The response for tracks incudes the start date and time of the asset moving, the total distance moved (in meters), and the total time that the asset was moving (in seconds).
\nIdle periods are detections of when the asset is at rest/stopped. The asset must be idle for at least as many seconds as the defined min_idle_time to consider the asset to be stopped and to consider the prior track to have ended.
\ntracks
\nidle
\n{\n \"tracks\": \n [\n {\"Date\": \"5/30/2021 18:28:39\", \"Dist\": 1.2008489725416527, \"TimeCounter\": 1},\n {\"Date\": \"5/30/2021 18:28:50\", Dist\": 30.835966982515377, \"TimeCounter\": 17},\n {\"Date\": \"5/30/2021 18:29:16\", \"Dist\": 33.574318610290405, \"TimeCounter\": 13}\n ],\n \"idle\": \n [\n {\"Date\": \"5/30/2021 18:28:30\", \"TimeCounter\": 9}, \n {\"Date\": \"5/30/2021 18:28:40\", \"TimeCounter\": 10}, \n {\"Date\":\"5/30/2021 18:29:07\", \"TimeCounter\": 9}\n ]\n}\n\nEndpoint to take in raw GPS coordinates and calculate insights around individual tracks.
\nThe endpoint calculates individual tracks and stops within the list of GPS coordinates and returns a summary of statistics for each track.
\nIn addition to the summary, the endpoint will also return the list of individual GPS coordinates that are part of the calculated track.
\nList of aggregated tracks and detected stops from the GPS coordinates.
\nA track consists of the path of a moving asset until it comes to rest. The response for tracks incudes the start date and time of the asset moving, the total distance moved (in meters), and the total time that the asset was moving (in seconds).
\nIn addition to the distance and time, the track reponse for the detailed endpoint includes the individual GPS coordinates along the track for mapping in third-party systems.
\nIdle periods are detections of when the asset is at rest/stopped. The asset must be idle for at least as many seconds as the defined min_idle_time to consider the asset to be stopped and to consider the prior track to have ended.
\ntracks
\nidle
\n{\n \"tracks\": \n [\n {\"Date\": \"5/30/2021 18:28:39\", \"Dist\": 1.2008489725416527, \"TimeCounter\": 1, \n \"track_coordinates\": \n [{\"Date\":\"5/30/2021 18:39\", \"Lat\": 41.518113, \"Long\": -87.390497}]}, \n {\"Date\": \"5/30/2021 18:28:50\", \"Dist\": 30.835966982515377,\"TimeCounter\": 17, \n \"track_coordinates\": \n [{\"Date\": \"5/30/2021 18:28:50\", \"Lat\": 41.518, \"Long\": -87.390539}, \n {\"Date\":\"5/30/2021 18:28:51\", \"Lat\": 41.518, \"Long\": -87.390518}, \n {\"Date\": \"5/30/2021 18:28:52\", \"Lat\": 41.517999, \"Long\":-87.390517}, \n {\"Date\": \"5/30/2021 18:28:53\", \"Lat\": 41.517996, \"Long\": -87.390515}\n ]\n ...\n ...\n ...\n ], \n \"idle\": \n [\n {\"Date\": \"5/30/2021 18:28:39\", \"TimeCounter\": 9, \"Lat\":41.518122, \"Long\": -87.390505}, \n {\"Date\": \"5/30/2021 18:28:50\", \"TimeCounter\": 10, \"Lat\": 41.518113, \"Long\": -87.390497},\n {\"Date\": \"5/30/2021 18:29:16\", \"TimeCounter\": 9, \"Lat\": 41.517961, \"Long\": -87.39053}\n ]\n}\n\nA collection of API endpoints designed to instantly analyze and provide insights on GPS coordinates.
\nGo beyond the simple location of your assets and gain valuable insights into how they are moving.
\n","_postman_id":"ce56009a-b071-4011-9d6a-34878fec6557"},{"name":"Data Aggregation","item":[{"name":"Describe Data","id":"4321816d-45f0-4a12-a47d-934ea2f85300","protocolProfileBehavior":{"disableBodyPruning":true,"disabledSystemHeaders":{"content-length":true,"content-type":true,"host":true,"accept":true,"connection":true,"accept-encoding":true}},"request":{"method":"POST","header":[{"key":"Content-Type","value":"application/json","type":"default"}],"body":{"mode":"raw","raw":"{\r\n \"Access_Key\":\"Endpoint to calculate and provide a statistical summary on the numerical columns in your dataset.
\nThe endpoint will take in your data and return statistics for each column name provided in the input. The column name will be the key in the response object with each column returning the following statistics for its values:
\nEndpoint to calculate the number of unique values for each column in your dataset.
\nThe endpoint takes in a list of objects where each object is a JSON representation of a single row in your dataset. For all columns containing fewer than 100 unique values, the endpoint will calculate and report the number of occurrences of each unique value in the column.
\nThe response consists of a list of JSON objects with the column name as the key followed by a dictionary of each unique value and the number of times that value occurs in the column.
\nThe response consists of a JSON object. Each column in the inputted data set will be a key in the response while the value of the key will consist of an object listing each unique value in that column and the number of occurrences for that value within the column of the dataset.
\nA collection of API endpoints to aggregate a user's data by groups or to provide statistical analysis on the dataset as a whole.
\n","_postman_id":"810ed5d3-f02a-4e00-bbd3-119fb7d1a667"},{"name":"Data Transformation","item":[{"name":"Count Missing Values","id":"3fd18376-82d5-4256-b9e9-67fc477138d9","protocolProfileBehavior":{"disableBodyPruning":true,"disabledSystemHeaders":{"content-length":true,"host":true,"accept":true,"accept-encoding":true,"connection":true}},"request":{"method":"POST","header":[{"key":"Content-Type","value":"application/json","type":"default"}],"body":{"mode":"raw","raw":"{\r\n \"Access_Key\":\"Endpoint to take in a dataset and count the number of missing values in each column.
\nThe endpoint takes in a list of objects where each object is a JSON representation of a single row in your dataset. For all columns containing fewer than 100 unique values, the endpoint will calculate and report the number of occurrences of each unique value in the column.
\nThe response consists JSON object where each column name from the input data is a key in the responses with the associated value being the number of missing values for that column in the dataset.
\nEndpoint to take in a dataset and fill in missing values.
\nOptional methods for filling missing values are forward fill (last included value), back fill (next included value), fill with the median value of the column, or fill with zero.
\nThe endpoint takes in a list of objects where each object is a JSON representation of a single row in your dataset. For all columns containing fewer than 100 unique values, the endpoint will calculate and report the number of occurrences of each unique value in the column.
\nThe response consists of the Data from the input payload with missing values filled in based on the method selected
\nEndpoint to take in a dataset and fill in missing values with a specified value.
\nThe endpoint takes in a list of objects where each object is a JSON representation of a single row in your dataset. For all columns containing fewer than 100 unique values, the endpoint will calculate and report the number of occurrences of each unique value in the column.
\nThe response consists of the Data from the input payload with missing values filled in with the value specified.
\nEndpoint to take in a dataset of values
\nThe endpoint takes in a list of objects where each object is a JSON representation of a single row with a single value in your dataset. Based on the lag values set in the Lags parameter, the endpoint will find the value X number of rows back in the dataset and add it to a new column on the current row.
\nFor example, if your data is recorded every minute and the lags provided are 5,15,30 the endpoint will find the value from 5 minutes prior, 15 minutes prior, and 30 minutes prior on a rolling basis and add them as new columns for each row.
\nThe response consists of the Data from the input payload plus columns for each of the lag periods that you set in the parameters
\nSeries of endpoints to perform common transformations on your dataset.
\n","_postman_id":"ee808a5d-3ee1-4bb8-94aa-0d772f69ac7a"}],"id":"3e6129a7-c93f-4ce0-a3f0-c7f474455a79","description":"Elipsa provides a series of endpoints to perform common data science and advanced analytic techniques in a single endpoint. The Advanced Analytics endpoints enable users to easily add advanced analytics into their workflows without writing custom code.
\n","_postman_id":"3e6129a7-c93f-4ce0-a3f0-c7f474455a79"},{"name":"EDGE ANALYTICS","item":[{"name":"Edge Administration","item":[{"name":"Set Access Key on Edge","id":"d02dff27-50cf-4f23-b26e-bef3c3cb1344","protocolProfileBehavior":{"disableBodyPruning":true},"request":{"method":"POST","header":[{"key":"Content-Type","value":"application/json","type":"default"}],"body":{"mode":"raw","raw":"{\r\n \"Access_Key\":\"Endpoint to set the user's Access Key on the Edge Device.
\nAll edge devices need to have a valid Access Key configured on the device for endpoints to function
\n","urlObject":{"port":"5000","path":["api","v1.0","access_key","set"],"host":["localhost"],"query":[],"variable":[]}},"response":[],"_postman_id":"d02dff27-50cf-4f23-b26e-bef3c3cb1344"},{"name":"Set Edge ID","id":"8a349e3a-6082-4fda-8428-30e195a3ee4e","protocolProfileBehavior":{"disableBodyPruning":true,"disabledSystemHeaders":{"content-type":true,"content-length":true,"host":true,"user-agent":true,"accept":true,"accept-encoding":true,"connection":true}},"request":{"method":"POST","header":[{"key":"Content-Type","value":"application/json","type":"default"}],"body":{"mode":"raw","raw":"{\r\n \"Edge_ID\":\"Endpoint to set the Edge ID for the edge device.
\nAll edge devices need to have a valid Edge ID configured on the device for endpoints to function
\n","urlObject":{"port":"5000","path":["api","v1.0","edge_id","set"],"host":["localhost"],"query":[],"variable":[]}},"response":[],"_postman_id":"8a349e3a-6082-4fda-8428-30e195a3ee4e"},{"name":"Get Edge Info","id":"9621672d-9782-4e05-8430-baf97feff09f","protocolProfileBehavior":{"disableBodyPruning":true},"request":{"method":"GET","header":[],"url":"localhost:5000/edge_info","description":"Endpoint to retrieve info of the Elipsa Edge deployment such as the version number, Access Key, and Edge ID configured on the device.
\n","urlObject":{"port":"5000","path":["edge_info"],"host":["localhost"],"query":[],"variable":[]}},"response":[],"_postman_id":"9621672d-9782-4e05-8430-baf97feff09f"},{"name":"Backup Edge Models","id":"5a56b3b8-57aa-4301-9feb-3d077f2afe1b","protocolProfileBehavior":{"disableBodyPruning":true},"request":{"method":"POST","header":[{"key":"Content-Type","value":"application/json","type":"default"}],"body":{"mode":"raw","raw":"","options":{"raw":{"language":"json"}}},"url":"localhost:5000/api/v1.0/support/backup","description":"Endpoint to run a backup of the models and templates deployed to the Edge Device.
\nThe backup endpoint allows users to download a zip file of models, templates, and configuration files to move to a new or upgrades Edge Deployment.
\n","urlObject":{"port":"5000","path":["api","v1.0","support","backup"],"host":["localhost"],"query":[],"variable":[]}},"response":[],"_postman_id":"5a56b3b8-57aa-4301-9feb-3d077f2afe1b"},{"name":"Restore Edge Models","id":"c9c4bc7f-947c-4490-adc6-db40e7a7267b","protocolProfileBehavior":{"disableBodyPruning":true},"request":{"method":"POST","header":[],"url":"localhost:5000/api/v1.0/support/restore","description":"Endpoint to restore the models and templates on an Edge Device.
\nThe restore endpoint allows users to take the zip from from the backup endpoint to move files to a new or upgrades Edge Deployment.
\n","urlObject":{"port":"5000","path":["api","v1.0","support","restore"],"host":["localhost"],"query":[],"variable":[]}},"response":[],"_postman_id":"c9c4bc7f-947c-4490-adc6-db40e7a7267b"},{"name":"Check in with Cloud","id":"e073dd8b-f2b6-451f-bfa9-5f4c8d3cd00d","protocolProfileBehavior":{"disableBodyPruning":true},"request":{"method":"GET","header":[],"url":"http://localhost:5000/api/v1.0/cloud/checkpoint/set","description":"Endpoint for the Elipsa Edge Device to check in with the Elipsa Cloud.
\nAt a minimum, Edge devices must check in with the Elipsa cloud once every 7 days for Edge functionality to operate correctly.
\n","urlObject":{"protocol":"http","port":"5000","path":["api","v1.0","cloud","checkpoint","set"],"host":["localhost"],"query":[],"variable":[]}},"response":[],"_postman_id":"e073dd8b-f2b6-451f-bfa9-5f4c8d3cd00d"}],"id":"ffe22738-b482-4348-aa30-724fe024190f","description":"A collection of endpoints to perform administrative tasks ensuring that edge devices are able to function properly.
\n","_postman_id":"ffe22738-b482-4348-aa30-724fe024190f"},{"name":"Self-Training Deployment Templates","item":[{"name":"List Deployed Templates","id":"52e5153f-3595-4f2d-8563-8e929a095654","protocolProfileBehavior":{"disableBodyPruning":true},"request":{"method":"GET","header":[],"body":{"mode":"raw","raw":"","options":{"raw":{"language":"json"}}},"url":"localhost:5000/api/v1.0/deployments/model/templated/list","description":"Endpoint to return a list of all Deployment Templates on the Edge Device.
\n","urlObject":{"port":"5000","path":["api","v1.0","deployments","model","templated","list"],"host":["localhost"],"query":[],"variable":[]}},"response":[],"_postman_id":"52e5153f-3595-4f2d-8563-8e929a095654"},{"name":"List Templated Models","id":"50ae859f-a0fb-4735-9586-eeddb586c027","protocolProfileBehavior":{"disableBodyPruning":true},"request":{"method":"GET","header":[],"url":"localhost:5000/api/v1.0/deployments/model/list","description":"Endpoint to return a list of all models on the Edge Device that are associated with a Deployment Template.
\n","urlObject":{"port":"5000","path":["api","v1.0","deployments","model","list"],"host":["localhost"],"query":[],"variable":[]}},"response":[],"_postman_id":"50ae859f-a0fb-4735-9586-eeddb586c027"},{"name":"Create Deployment Template","id":"eb38644d-2e99-48db-8e5a-55ceddf33b63","protocolProfileBehavior":{"disableBodyPruning":true,"disabledSystemHeaders":{"content-type":true,"content-length":true,"host":true,"user-agent":true,"accept":true,"accept-encoding":true,"connection":true}},"request":{"method":"POST","header":[{"key":"Content-Type","value":"application/json","type":"default"}],"body":{"mode":"raw","raw":"{\r\n \"Name\":\"Endpoint to create a new Deployment Template on the Edge Device.
\nEndpoint to stream new telemetry data to a given template (specifying the template's API Key in the url).
\nThe router utilizes the unique ID parameter to automate the AI model creation, deployment, and inference.
\nSimply connecting a machine to the template for future data streaming enables self-training no-touch outlier detection.
\nElipsa Edge Analytics puts the power of advanced analytics and self-training AI directly on your edge device.
\nFrom Predictive Maintenance to Intelligent Asset Monitoring, users can now take advantage of the same easy to use APIs from Elipsa that are available in the cloud.
\nElipsa Edge Analytics requires an edge license. To learn more contact support@elipsa.ai.
\n","_postman_id":"d87769c4-af2b-4a68-8cfe-a02d408f9c2f"}]}