How to Create a Time Study
Step 1: Capture and Upload Process Videos
Provide the agent with clear videos of the process you want to study.
The agent automatically analyzes the footage and produces a complete time study in just a few minutes.
Step 2: Attach Supporting Line Documentation
Attach all relevant documents associated with the line so the analysis is grounded in your actual constraints and standards. This can include:
Work instructions
Specification sheets
Downtime logs
Bills of materials (BOMs)
Step 3: Configure Line Constraints and Business Rules
Tell the agent what you already know about your operation so results are tailored to your environment. For example, specify:
Equipment that cannot be moved
The internal names your team uses for parts
Any cap on spending for improvements
Tests or quality checks that must remain in place
The agent will build its recommendations around these constraints.
Step 4: Start a New or Existing Variance Analysis
Begin analyzing performance and variation by either:
Starting a new variance analysis, or
Adding the new footage and data to an existing analysis
Step 5: Protect Operator Privacy (Optional)
Before generating your analysis, you can enable privacy controls to blur operators’ faces in the footage so that personal identities are not visible in the resulting study.
Step 6: Review the Automatically Generated Timeline
The agent breaks each video into discrete work elements and lays them out on a continuous, gapless timeline that represents the full process.
Each element in the study is:
Timestamped
Classified as value-added or non-value-added
Tagged with the relevant waste category
Step 7: Edit and Refine Work Elements
Use the interactive timeline to fine-tune the study while keeping the underlying footage intact.
You can:
Reclassify any work element (e.g., change value-added vs. non-value-added)
Drag boundaries to adjust where elements start and end
Add elements that the agent may have missed
All edits are saved and shared across your workspace so everyone works from the same, updated study.
Step 8: Prioritize Improvements with Ranked Analysis
Use the analysis view to understand where you will get the most impact.
The agent ranks improvement opportunities by how much time they consume. Each finding includes:
A specific timestamp in the footage
A concrete recommendation
Alignment with the constraints and rules you configured earlier
Step 9: Analyze Cycles Across Long and Multiple Videos
Let the agent detect and group every cycle, even within long recordings and across multiple videos.
The resulting time study can span:
Multiple days
Different shifts
Different operators
Compare cycles and work elements side by side to see:
Median performance
Average cycle times
Outliers and variation drivers
Step 10: Model the Line and Identify Bottlenecks
Bring multiple time studies together so Squint can model the full line. The Yamazumi chart stacks each station’s workload against takt time, making the bottleneck clearly visible.
Below the Yamazumi, review:
Cycle time by station
Balance efficiency
Every issue the agent flagged at each station, including safety concerns
The analysis spans the entire line, highlighting:
Shared sources of waste
Handoffs between stations
Work that can be moved off the bottleneck
Each recommendation is prioritized and shows expected savings per unit, backed by timestamped video evidence.
Step 11: Rebalance the Line and Run Simulations
Use the rebalancing tools to test new configurations before making physical changes.
Specify your targets, such as:
Desired number of stations
Target cycle time
Then run a simulation. Watch work move through every element on the line and see precisely where it backs up.
Click any element in the simulation to inspect what’s happening at that point. For example, you might see a station running over takt because an operator is idle, waiting, or chatting.
Run the same simulation again on a rebalanced configuration to confirm whether the new balance holds under realistic conditions—before you make any changes on the floor.
