Tutorial: Stamping & Punch Presses
Perfect stroke counting for stamping presses using Discrete w/o Merge algorithm combined with SenseAi downtime filter.
Overview
Stamping and punch presses generate distinct vibration spikes with each stroke, making them perfect for precise counting using the Discrete Analysis w/o Merge algorithm. This tutorial shows you how to achieve accurate stroke counting while eliminating micro-stop noise.
What you'll learn:
- Why Discrete w/o Merge is perfect for presses
- How to combine algorithm (Parts List) with device downtime filter (SenseAi)
- Achieving clean data with perfect stroke counts
- Handling high-speed vs. low-speed presses
Time required: 30-45 minutes for complete setup
Why Discrete w/o Merge for Stamping Presses?
Stamping presses create distinct, countable events (strokes) that should each be tracked:
✅ Discrete strokes: Each press stroke is a separate, detectable event
✅ High frequency: Can run 30-120 strokes per minute
✅ No merging needed: Every stroke should count individually
✅ Brief micro-stops: Small pauses between strokes should be ignored
The secret: Combine Discrete w/o Merge algorithm (in Parts List) with Downtime Filter setting (on the SenseAi device itself) for perfect results.
The Two-Layer Filtering Approach
Understanding this is key to success:
| Layer | Where Configured | What It Does |
|---|---|---|
| Algorithm | Parts List (operation settings) | Discrete w/o Merge counts every detected vibration spike as a stroke |
| Device Downtime Filter | Devices Tab (SenseAi sensor settings) | Converts all downtimes under X minutes to uptime |
Combined result: Every stroke is counted, but only meaningful downtimes (longer than X minutes) are captured.
Step-by-Step Setup
Step 1: Install and Calibrate SenseAi
Mount SenseAi on the press frame
Attach close to the ram or drive mechanism where vibration is strongest. Ensure metal-to-metal contact for best sensitivity.
Power and connect
Connect USB-C power and pair to WiFi using the IoTFlows mobile app.
Run test strokes
Manually cycle the press 5-10 times and verify SenseAi is detecting each stroke in the Devices Tab (look for activity spikes).
See the full SenseAi Installation Guide for detailed instructions.
Step 2: Measure Your Stroke Rate
Run the press normally
Start production and let the machine run for 5 minutes
Count actual strokes
Use the press counter (if available) or manually count strokes for 1 minute. Record the number.
Example: 60 strokes in 1 minute = 60 SPM (strokes per minute)
Calculate cycle time
Cycle time = 60 seconds ÷ strokes per minute
Example: 60 seconds ÷ 60 SPM = 1.0 second per stroke
Note typical pauses
Observe any brief pauses between strokes (usually microseconds to 1-2 seconds max).
Step 3: Create Operation in Parts List
Navigate to Parts List
Go to Production Tab → Parts List
Click Add Part
Click + Add Part
Fill in operation details
- Operation Name: e.g., "STAMP-BRACKET-100T"
- Part/Material: e.g., "Steel Bracket"
- Description: e.g., "100-ton stamping press producing brackets"
Select Discrete Analysis w/o Merge
From the Algorithm dropdown, choose Discrete Analysis w/o Merge
Enter cycle time
Input the cycle time you calculated (e.g., 00:00:01 for 1 second per stroke)
Set Quantity Per Cycle
- Most common: 1 (one stroke = one part)
- If multiple parts per stroke (e.g., progressive die), adjust accordingly
Save the operation
Click Save
Do NOT set a Downtime Filter in Parts List for this algorithm. Leave it at 0% or empty. You'll configure downtime filtering at the device level (next step).
Step 4: Configure SenseAi Device Downtime Filter
This is the critical step that makes everything work perfectly.
Navigate to Devices Tab
Go to Devices Tab
Select your SenseAi sensor
Click on the SenseAi device attached to the stamping press
Open device settings
Click Edit Sensor or similar settings button
Set Downtime Filter (device level)
Find the Downtime Filter setting for the device and set it to 10 minutes (or 5-15 minutes depending on your preference)
Understand what this does
- Algorithm counts every stroke/spike
- Any downtime under 10 minutes is converted to uptime
- Only stops ≥10 minutes are recorded as downtime
Result: Perfect stroke counting with clean downtime data (no micro-stops cluttering your reports)
Save changes
Click Save or Update
Device-level downtime filter vs. Parts List downtime filter are different!
- Device level (what you just configured): Filters out ALL short downtimes before any algorithm sees them
- Parts List level: Works within the algorithm calculation (not used for Discrete w/o Merge)
For stamping presses, always use device-level filtering.
Step 5: Assign Operation to Machine
Go to Assets Tab
Navigate to Assets Tab → Overview
Click on the stamping press
Open the machine detail view
Click Auto-Detect
Find and click the Auto-Detect button
Select your operation
Choose the stamping operation you just created (e.g., "STAMP-BRACKET-100T")
Confirm selection
The machine will now count every stroke while filtering short downtimes
Step 6: Validate and Monitor
Run for 1-2 hours
Let the press run normally
Compare counts
Check IoTFlows count vs. press counter (if available):
Validation:
- Press runs 60 strokes/min for 60 minutes
- Expected: 3,600 strokes
- IoTFlows should show 3,600 parts (±2-3%)
Review downtime events
Go to Assets Tab → Downtimes:
- Should only see stops ≥10 minutes (material loading, die changes, lunch breaks)
- Should NOT see brief pauses between strokes
Adjust device downtime filter if needed
- Too many micro-stops showing: Increase to 15 minutes
- Missing real downtimes: Decrease to 5 minutes
Common Issues & Solutions
Recommended Downtime Filter Settings
| Press Type | Strokes/Min | Device Downtime Filter | Why |
|---|---|---|---|
| High-speed stamping | 60-120 SPM | 10-15 minutes | Eliminates brief pauses; captures only major stops |
| Medium-speed punching | 30-60 SPM | 5-10 minutes | Balances detail with noise reduction |
| Low-speed forming | 10-30 SPM | 3-5 minutes | More sensitivity needed for slower operations |
| Manual feed presses | Variable | 2-5 minutes | Captures operator-related pauses |
Start with 10 minutes and adjust based on your downtime report. If you see too many small events, increase it. If you're missing real stops, decrease it.
Best Practices for Stamping Presses
1. Progressive Dies vs. Single-Hit Dies
Progressive dies (multiple operations per stroke):
- Each stroke completes one part through multiple stages
- Set Quantity Per Cycle = 1 (one stroke = one completed part)
- Focus on stroke count, not individual operations
Single-hit dies (one operation per stroke):
- Each stroke = one operation
- Set Quantity Per Cycle = 1
2. Die Changes and Setup
Train operators to classify downtime during die changes:
- "Die Change" category for planned changeovers
- "Setup/Adjustment" for fine-tuning after installation
Typical die change: 15-60 minutes (will be captured with 10-min filter)
3. Material Loading
If coil loading takes 5-10 minutes:
- Ensure device downtime filter ≥5 minutes to capture these events
- Create auto-classification rule for "Material Loading" based on duration patterns
4. Set Realistic OEE Goals
Stamping operations typically achieve:
- Manual feed: 50-65% OEE
- Auto feed (coil): 65-80% OEE
- Lights-out automation: 80-90% OEE
Start 5-10% above your current baseline.
5. Monitor Stroke Rate Trends
Use Historical Production to track strokes over time:
- Declining stroke rate → tooling wear, press degradation
- Increasing downtime → material issues, quality problems
Real-World Example
Company: Precision Metal Stamping Co. Machine: 75-ton mechanical press Part: Steel automotive brackets Stroke rate: 45 strokes per minute
Configuration:
- Algorithm (Parts List): Discrete Analysis w/o Merge
- Cycle Time:
00:00:01.33(60 ÷ 45 SPM) - Quantity Per Cycle: 1
- Device Downtime Filter (SenseAi): 10 minutes
Results:
- Before IoTFlows: Unknown press utilization, frequent short stops unnoticed
- Week 1: 62% utilization, ~18 downtime events/day (all ≥10 minutes)
- Month 1: Identified material loading as #1 downtime (35% of total)
- Month 3: Improved material staging process, increased utilization to 74%
- ROI: 12% production increase = $38K annual value
Next Steps
Monitor Strokes in Real-Time
View live press activity in Assets Tab
Track Production Goals
Monitor shift stroke targets
Analyze Downtime Patterns
Identify improvement opportunities
Have questions about press monitoring setup? Contact support@iotflows.com for personalized guidance!

