The SPEDE USB Camera solution uses simple off-the-shelf cameras at production lines and pack stations to error-proof packing operations by verifying and counting parts as they're packed into containers.
- Prevents short-ships
- Putting the wrong part in a container
- Mis-identifying parts from overseas at re-pack
Uses Simple USB Cameras
- Cameras cost as little as $75
- Cameras have no intelligence, so there is no programming or configuration of the camera needed
- Cameras simply capture an image for SPEDE Vision logic to analyze
- No PLC is needed. SPEDE uses image capture and pixel analysis on a video frame-by-frame basis
- Cameras can integrate with Mitsubishi and Koyo PLCs
- All image processing is performed on a PC workstation connected to the camera(s)
- All Part Number targeting data (master image, size, shape, etc.) is stored in a common SQL database that all lines can access
- No local master image files; any SPEDE line can use the same
master images
- No need to maintain local master image files on multiple PC workstations
- Camera can be “trained” to recognize a part at your office desktop PC – No need to perform this task at the line
- Multiple cameras can be used to capture multiple angles without excessive incremental costs
- 1080p cameras at 16:9 aspect provide much greater field of view
- Cameras can be easily swapped out by maintenance
SPEDE Vision logic is "trained" to recognize a part, and stores the part master image in an SQL database on a PC workstation, which all lines can access.
SPEDE Vision uses simple USB cameras for:
- Part verification - compares image of part to master part image; shunts incorrect part downstream
- Color detection - compares colors present to master image; shunts incorrect part downstream
- Counting parts - increments part count if part is correct
- OCR - reading 1D, 2D barcodes and characters on parts and labels
- Controlling the packing sequence of dunnage layers
- Verifying all dunnage layers are filled correctly, with video proof
- Verifying pack count is correct before triggering a container label