Tutorial: Your First Machine Setup

Select a location on the machine where vibration is strongest:

• Close to the motor, hydraulic pump, or spindle
• Metal-to-metal contact (avoid rubber, plastic, or painted surfaces)
• Accessible for USB-C power cable

Attach SenseAi using the magnetic base or adhesive mount:

• Ensure firm contact with metal surface
• Verify sensor won't interfere with machine operation
• Position LED indicator where it's visible

Plug in the USB-C power cable to a nearby outlet or USB power adapter

SenseAi LED should light up (blue or green depending on status)

• iOS: Download from App Store
• Android: Download from Google Play Store

Use your IoTFlows account credentials

Follow the in-app instructions to pair SenseAi to your WiFi network

Give it a descriptive name (e.g., "Molding-Machine-101-SenseAi")

Check that the device shows "Connected" status in the app and web platform

In the IoTFlows web platform, go to Devices Tab

Click on the device you just added

Start the calibration process

Let the machine run for 5-10 production cycles while calibration is active

After calibration completes, stop the machine briefly

SenseAi automatically sets the sensitivity threshold based on the calibration run. Verify it captured machine activity correctly.

Go to Production Tab → Parts List

Create a new operation

• Operation Name: Descriptive name (e.g., "INJECTION-WIDGET-A")
• Part/Material: Part being produced (e.g., "Plastic Widget")
• Description: Optional notes about the operation

Choose the algorithm based on your machine type:

• Continuous: For injection molding, consistent cycle times
• Discrete: For CNC machines, variable cycle times
• Discrete w/o Merge: For stamping presses, stroke counting

See Choosing the Right Algorithm for detailed guidance.

Measure your machine's actual cycle time:

• Time 10 production cycles with a stopwatch
• Calculate average (e.g., 22 seconds per part)
• Enter in format 00:00:22

• Single part per cycle: 1
• Multi-cavity mold (4 parts): 4
• Progressive die producing 2 parts: 2

For Continuous algorithm:

• Set to 200% as a starting point
• This converts downtimes under 44 seconds (22s × 2) to uptime

Click Save or Add Operation

Go to Assets Tab → Overview

Locate the machine in the assets list

Open the machine detail view

This opens the operation selection modal

Choose the operation you just created (e.g., "INJECTION-WIDGET-A")

The machine now starts tracking production using your configured algorithm and cycle time

In Assets Tab → Overview, verify the machine shows "Running" status when active

Go to Production Tab → Shift Production and find your machine's production count

Calculate expected parts:

• Runtime: 60 minutes = 3,600 seconds
• Cycle time: 22 seconds
• Expected: 3,600 ÷ 22 ≈ 163 parts
• 4-cavity mold: 163 × 4 = 652 parts

Your IoTFlows count should be within ±5-10% of this calculation.

Go to Assets Tab → Downtimes and verify:

• Real stoppages are captured (material loading, breaks, setup)
• Brief pauses are NOT creating clutter (if they are, increase downtime filter)

• Count too high: Increase downtime filter or verify cycle time
• Count too low: Decrease downtime filter or check sensor calibration
• Way off: Wrong algorithm selection or incorrect cycle time

Run for at least 2-3 days to establish a baseline utilization

Go to Assets Tab and note your current utilization % (e.g., 62%)

Click on the utilization percentage to open the goal modal. Set a goal 5-10% above current baseline:

• Current: 62%
• Goal: 68-70%

Track daily utilization and work to close the gap to your goal

Go to Organization Settings → Auto-Downtime Classification

Add common downtime reasons:

• Material Shortage
• Setup/Changeover
• Maintenance
• Quality Issue
• Operator Break

Show operators how to classify downtimes:

• Use mobile app to classify in real-time
• Or classify from web platform at end of shift

After a week, go to Assets Tab → Downtimes and view Pareto chart to identify top losses