Self-Checkout: The Shift from Weighing Scales to Visual AI
Sainsbury’s has taken the lead with a recent announcement which talks about rolling out Visual AI in their network of stores across the UK (see news). Other retail companies are in advanced stages of evaluating the addition of Visual AI into their mix of store security solutions.
The Retail Security Landscape
Retail companies across the world are evaluating and implementing various technological solutions to contain rising shrinkage, which has resulted in billions of dollars in losses for the industry. Traditional measures such as deploying security personnel, educating store staff, and increasing CCTV coverage have all only served as security deterrence.
Self-Checkout is Fraught with Loopholes
Self-checkout (SCO) was expected to reduce the burden of monitoring and enable friction-free shopping for customers. As one of the benefits of SCO is that it reduces manpower costs for the retail company, SCO is witnessing largescale deployments. However, with billions of dollars lost in shrinkage, one-fifth of which is being attributed to SCO, retail companies are now waking up to the reality.
Here are a few common types of frauds that typically occur at a SCO counter:
| Cause of Loss | Customer Behaviour |
|---|---|
| Incorrect Scans | The customer deliberately weighs lower priced products. |
| Product Switching | The customer weighs a different product to avoid paying for the actual price of the product. |
| Missed Scans | The customer fails to scan items from the cart or moves items over the scanner in such a way that they are not scanned. |
| Non-Payment | The customer walks out without finalizing the payment for the items scanned. |
Additional Workload for Store Staff
To cover the risks associated with incorrect scans, retail companies have evaluated the use of weighing scales at the SCO counters. However, to implement weight-based verification, every SKU in the inventory needs to be scanned and the weight of each item needs to be accurately captured and integrated into the SCO. Any new packaging or a brand-new item can introduce additional challenges.
Weighing scales also need to be calibrated to ensure accuracy. Additionally, correlating an item solely based on its weight and its corresponding barcode may not prevent a customer from scanning the incorrect barcode. Scanning of cheaper items that are equal in weight when compared with the actual item, cannot be detected.
For instance, swapping a bunch of bananas with a nice bottle of Chardonnay cannot be detected by this method. Also, in the hustle and bustle of packing, if the customer places any personal item, such as a purse, on the weighing scale, it can trigger erroneous alerts. For all these reasons, retail companies are considering removing scales from the counters.
Visual AI Gaining Popularity with Retail Companies
Of the many technical approaches available such as analytics, a Visual AI solution using the existing infrastructure appears to be the apt choice. The staff can get alerted when a potential SCO loss is detected, enabling them to quietly intervene and nudge the customer to scan the items again, this time with the correct product code.
All of this can be done without escalating the situation and ensuring that other shoppers are not alarmed. A Visual AI solution, which can also provide video footage of events, can minimize customer disputes at the SCO counters.
SAI Group provides Visual AI, data analytics, intuitive dashboards, and robust reporting capabilities to three of the top retail companies in the UK. Contact us to enhance your retail organization’s loss prevention capabilities.
About SAI
As a leader in computer vision technology, SAI Group delivers cutting-edge, multi-modal AI solutions into retail environments. Using a unique platform approach, its technology uses existing camera systems to target losses, increase store safety, and underpin operational efficiencies.
All solutions are built from the ground up to ensure the highest levels of security and data protection, respecting the privacy expectations of the public and operating to stringent ethical standards while delivering substantial value to our clients. Globally, SAI monitors millions of transactions per day, protecting the revenues from tens of millions of product sales and hundreds of millions of customer interactions. Its models also accurately identify anti-social behavior, aggression and violence, helping to de-escalate situations with real-time interfaces to security officers and operations centers.