Introduction
Overview
This tutorial introduces Sift through a hands-on, end-to-end workflow using real-world telemetry. You will work with a preprocessed dataset from the Mars Environmental Monitoring Station (REMS), an instrument package aboard NASA’s Curiosity rover. The preprocessed dataset contains one year of environmental sensor readings collected on the Martian surface, including temperature, pressure, and ultraviolet radiation levels.
You will begin by importing and visualizing the dataset to explore environmental patterns and spot potential anomalies. After identifying a sharp temperature drop, you’ll define a condition to automatically detect similar behavior in the future. You’ll then create a summary view that highlights when and where this condition occurs across the dataset. Finally, you’ll generate a new signal by combining temperature readings, making it easier to track changes over time. This mirrors a typical hardware telemetry workflow: importing and exploring data, identifying patterns, automating checks, and extracting insights for review and analysis.
Prerequisites
- Preprocessed dataset: env_telemetry_two_thousand_twenty_one.csv
- Access to Sift (either the Public Cloud or GovCloud deployment)
Preprocessed dataset
| earth_date_time | mars_date_time | sol_number | max_ground_temp(C) | min_ground_temp(C) | max_air_temp(C) | min_air_temp(C) | mean_pressure(Pa) | sunrise | sunset | UV_Radiation | weather |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 2021-12-20T00:00:00Z | Mars, Month 5 - LS 144 | Sol 3332 | -6 | -73 | 7 | -80 | 703 | 05:33 | 17:23 | moderate | Sunny |
Steps
Understand how Sift organizes telemetry data
Learn how Sift uses Assets, Channels, and Runs to structure time-series telemetry. These concepts will help you understand how signals are grouped, stored, and queried throughout the tutorial.
Review the preprocessed dataset used in this tutorial. You'll get familiar with the columns, units, and type of environmental signals collected by NASA’s Curiosity rover on Mars.
Import rover data and visualize it
Bring the preprocessed dataset into Sift by creating a new Asset and uploading the telemetry CSV file. Once uploaded, you'll explore temperature and other weather signals using Sift's visual charting tools.
Capture and detect an event in the data with a Rule
Use your observations to define a Rule that detects an unusual pattern in the data. This step shows how to codify engineering insight into reusable logic for automatic anomaly detection.
Generate a Report based on Rule results
Apply your Rule to a new dataset and generate a Report to see how Sift automatically flags any violations, linking detection to structured review.
Create a new signal using Calculated Channels
Define a new signal by combining existing telemetry using expressions. You'll create a Calculated Channel to track air temperature swings, demonstrating how Sift helps reveal new insights without modifying raw data.