Trends

Median

Median is a data reduction method that represents each time interval by its median value instead of using average or direct sampling. This helps preserve outlier robustness and provides a more stable trend line in noisy environments.

How It Works:

  • The time range is divided into fixed intervals (based on Max Points)
  • All values within each interval are collected
  • The median value (the middle value in the sorted list) is chosen to represent that segment

Industrial Use Case

Scenario: Monitoring vibration signals in rotating machinery

High-frequency vibration data is collected from a motor. These values often include noise spikes due to mechanical inconsistencies.

Problem:

  • Simple averaging may be distorted by outliers (e.g., a random shock)
  • Important trends can be lost in visual noise

Solution:
By using Median:

  • The trend line remains stable and less affected by occasional spikes
  • It offers robust smoothing, preserving the central tendency
  • Ideal for diagnostic dashboards or long-term monitoring of unstable signals