Biomechanical analysis often involves repeating the same steps each time new data is collected. Even when you standardize these steps, the workflow still depends on someone being available to initiate each stage. For every trial, you must locate your file, open it in Visual3D, run the correct pipeline, wait for processing to finish, save the output, open Sift, load in the results, and begin the next analysis. Repeating this process for dozens of files takes time, interrupts other work, and makes delays or inconsistencies in processing. Sift’s new Directory Watchers make it possible to automate biomechanics analysis by processing new files as soon as they are added to a folder and without user interaction.
So what if finishing data collection could also start the next stage of processing with no input required?
How Directory Watchers Automate Biomechanics Analysis
Directory Watchers fit any workflow with a repeatable sequence of processing steps. Directory Watchers help users:
- begin processing as soon as data collection is complete
- perform timely checks before a participant leaves the lab
- apply the same analysis consistently across subjects, sessions, or collection sites
- automate large-scale biomechanical analysis with no user intervention
Adding a C3D file to a watched folder is the conventional trigger event for automatically running the assigned Visual3D pipeline and creating a CMZ file in the chosen output folder.
Example: Kinematic Analysis Workflow
In this example, we add a treadmill-running C3D file to the watched Data folder. The watcher launches a Visual3D pipeline, which calculates the right and left ankle joint angles and saves the processed trial.
A second watcher detects the processed file and launches a Sift batch script. Sift extracts the right and left ankle-angles, normalizes signals to the gait cycle, runs Principal Component Analysis, and uses Local Outlier Factor to identify unusual curves.
The workflow follows:
Sift exports the final results and saves the analysis as a project. Then, you can review the original signals, principal-component scores, and detected outliers. This workflow allows initial results to become available while data collection continues. This is especially useful if you need to review data quality before a participant leaves the lab.
Customizing Your Automation
You can create and manage Directory Watchers directly through the Sift interface. Each watcher monitors one or more folders and launches its assigned Visual3D or Sift scripts when the watcher detects a supported file change. You can configure each watcher with:
- one or more monitored directories
- one or more Visual3D or Sift scripts
- a delay before processing begins
- optional recursive subfolder monitoring
- optional triggers when a user deletes a file
- an active or inactive state
When a watcher contains multiple scripts, you can adjust their order to control the execution sequence. The Sift Directory Watchers dialog gives users an easy-to-use method of customizing these sequences to match their needs exactly.
Directory Watchers also include options for more flexible workflows. You can monitor nested folder structures, respond to file deletions, and add a delay before processing begins. To simplify things even further, Watchers run from the Windows system tray where Sift can monitor folders without launching the interface.
Automate for Efficient Processing
Processing a large dataset? Watchers leverage the Command Line, bypassing the graphical interface.
The graphical interface continually updates plots, tables, queries, and analysis pages while an analysis runs. These updates add computational overhead, especially for large libraries. Command-line processing skips most of these updates and can complete the same analysis more quickly.
You still configure the workflow through the Directory Watchers dialog, while the scripts perform the computational work efficiently in the background.
Learn more about the performance and automation benefits in our post on Sift’s command-line functionality.
Learn More and Get Started
Directory Watchers turn repetitive processing steps into a workflow that starts on its own. This means less time spent opening files and waiting to start the next process, and more time reviewing the data, identifying issues, and focusing on the analysis.
Want to learn more about how you can automate biomechanics analysis? Check out our Directory Watcher Tutorial for a step-by-step guide to creating and activating watchers. You can also learn more about the Directory Watchers dialog and its settings in our feature overview.
Have questions or want to explore how this new feature can benefit your workflow? Contact us at in**@********on.ca to schedule a demo or discuss how Sift can help streamline your data processing workflows.