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