ScanWatch Retail Suite is an AI powered application to prevent SCO losses and improve customer experience.
The application offers a comprehensive security layer to address the absolute majority of SCO loss scenarios, which could not be addressed by traditional weight-based security platforms.
ScanWatch Retail Suite is hardware agnostic and can be installed on nearly any type of SCO hardware.
The platform is already protecting more than 5.000 SCOs, with top retail chains in Germany, Switzerland, Lithuania, and Poland relying on ScanWatch to enhance security, efficiency, and customer experience.
Item switch checks whether the scanned product image matches the one in the product database.
If a mismatch is detected, the system issues a warning, indicating the specific checkout station where it occurred.
Optionally, the warning can be displayed to the customer on the SCO screen (with a self-correction prompt). Store attendant can be notified as well.
This program aims to detect cases where the scanned barcode of a product does not match the product image.
Using computer vision technology, the program identifies the product and compares the scanned barcode with the correct one stored in the system. It identifies incidents when barcodes could be intentionally switched (a barcode from a cheaper item placed on an expensive product).
If a mismatch is detected, the system alerts store staff in real time, indicating the specific checkout station.
This program is designed to detect cases where scanning is simulated. Scan avoidance occurs when the scanning gesture is imitated, but the item is not actually scanned or added to the shopping list.
The program analyzes video streams in real time to determine whether scan avoidance may have occurred. This is especially useful at self-checkouts without security scales, as well as at regular checkouts where a cashier may be suspected of intentionally letting items pass through without scanning them.
Using computer vision technology, this program counts the number of items present at the self-checkout (within a system-defined and calibrated zone) and compares it with the number of scanned items.
This program is most useful at self-checkouts that do not have security scales.
This function checks whether all items are scanned and paid for.
It aims to prevent a scenario where a customer scans and pays only for some items, while leaving others in the shopping basket. After making the payment, the customer then proceeds to bag the unpaid items from the basket and leaves the checkout area.
Therefore, the program detects whether any items remain in the basket at the moment the customer initiates the payment process (by pressing the “Pay” button on the self-checkout screen).
If the basket is not empty, the program alerts store staff in real time, indicating the specific checkout and providing a photo of the basket area showing the remaining items.
This program monitors whether the scanning and payment process at the self-checkout is completed with an actual payment.
It analyzes transmitted video data. If a customer leaves the self-checkout area and remains absent for longer than a predefined number of seconds, the system alerts store staff in real time and provides two images of the customer (top-down and side views) to help identify the individual.
Picklist Assistant visually identifies the product without the need to select it from the SCO menu.
Identified item can pop-up as recommendation (‘Is this the correct item?’) Item can be automatically added to the receipt without customer confirmation.
Picklist Assistant significantly reduces the number of interventions and assistance requests, making purchasing bulk items faster and more user-friendly. By reducing unpackaged item scan time from 9 to 2 seconds, Picklist Assistant can increase SCO throughput by more than 30%.
