Retail shrinkage is a continuous problem and will continue to be a major issue affecting the UK. The 2017 British Retail Consortium Crime Survey reveals that UK shops lost more than £1.8 billion in a year via theft. The figure is damaging to the company’s finances, but it does not even include operational disruptions, expenditures of personnel welfare and insurance charges.
The typical industrial reaction to theft has been more CCTV, more security guards and more locks. This has generated relatively modest effects. The necessity to defend the business remains the same and critical, but the use of new technologies is what’s different. One of the most important developments in dual-purpose employee protection and theft prevention is the use of the Protex AI retail safety platform. Protex Technology is a safety platform that uses computer vision to assess safety in retail, detecting threats (safety hazards) and supporting loss prevention in real-time, rather than examining loss prevention film after a theft has occurred.
What Are the Limitations of Traditional Retail Security?
- Generally, the standard security systems of retail businesses have limits that cannot be addressed by making cosmetic improvements.
- By their very nature CCTV systems are retrospective. They understand what has occurred and do not understand what will happen. “By the time a retail security incident is reviewed, the loss has occurred and the opportunity is gone.”
- Humans don’t scale. A security worker cannot watch a dozen camera feeds simultaneously. Sustained attention tests reveal that attention for high quality watching deteriorates after only 20 minutes. No amount of additional manpower will fix this situation.
- Static deterrents are a learnable security technique. Retail criminal organizations take advantage of loopholes in static security measures resulting in a disproportionate loss of value.
- Inconsistent reporting of incidents Employees have a lot of reasons to not report safety and near miss accidents. So management does not know where the real hazards are.
- These examples all point to a similar security weakness. Retail security has historically been reactive. These systems respond to incidents and generate extra expenditures, rather than averting the losses.
What AI Safety Technology Actually Does
- When computer vision systems are built for retail, the systems focus on the present in a manner that traditional security monitoring approaches do not.
- Events are not logged for subsequent viewing in these systems. Instead the analysis is performed on live video streams to detect behavioral and environmental trends likely to precede a security problem.
- These systems highlight hazards and issues in real time, so that management and system stakeholders may take action to minimize risk.
- There is a lot more practical use than just theft prevention. There are slip and fall dangers, restricted emergency exits, stock risks and lone worker threats.
- Human monitors suffer from attention fatigue. AI does not. This is a big win for multi-site merchants. AI creates a safety baseline for each retail site, irrespective of the shop management or crew quality.
Broader Business Considerations
- When merchants examine AI safety, it’s frequently seen as a theft prevention tool. But the impact on loss prevention is far larger.
- “Often, the cost of accidents at work is assessed by the increased insurance. However, the accompanying expenses of management time, loss of business productivity and loss of circumstances might mean that claims can cost tens of thousands. Investing in AI safety measures up front saves a lot of money in the long term.
- AI safety is closely related to staff retention. Perceptions of safety at work are a cause why people leave their employment and retail is one of the biggest employee turnover jobs in the UK. A protected employment setting (even in the case of a safety occurrence) is maintained at a higher rate.
- Making AI safer is an investment that pays for itself in lower insurance and safety expenses.
Where Fast Adoption of Retail Safety AI is Most Common
The quickest adopters of AI safety tools across retailers have certain common qualities. They are likely to be in an area with considerable foot traffic and occurrences are statistically expected. They tend to run several sites where uniformity in safety requirements is difficult to maintain.” They prefer to operate huge formats where the expense of theft, injuries and even reputational issues are high.
AI safety technologies are most likely to be used by cash and carry operations such as big format grocery, large format convenience, large format fashion and large format wholesale. These are often the pilot operations to which the technologies are scaled.
What It Means for Retail Safety AI Adoption in the UK
The AI discussion in retail has mostly been about demand, personalization and inventory. Safety and loss prevention are already emerging as the operational use cases of safety AI where the highest advantages from the fastest and most readily quantifiable adoption are expected to be discovered.
Loss, injury claims, insurance costs and worker retention all provide a strong argument for safety AI solutions in a retail safety AI infrastructure investment for finance directors and operations officers.
Unless retail safety and risk management technology improve, the cost of retail crime will remain the same. But change is not only the observation of Camera systems. Real-time detection and alerting capabilities of active sensor technology are.
