The role of recognition technologies in creating safer stores for everyone

Crime in retail is not simply about shoplifting or petty theft. It's now a clear violence problem where threats with weapons, intimidation, and outright assault are used as tactics by crime groups and prolific offenders to steal products destined for black markets.

Aside from the abhorrent brazen and violent offending issues, crime in retail is also a challenge of scale, and is the highest volume crime type in many countries. 

It’s this scale that also makes it such a challenge for retailers and law enforcement. Simply put — they cannot chase everything. 

Retailers are therefore combatting both serious safety issues and high-loss challenges. In practice, that means they are spending billions on what’s known as ‘target hardening’.

These are all the physical and technical barriers that make it harder to steal such as one-way gates, cameras, guarding. Ultimately, we all pay for those costs when we shop. 

The question for those partnering with retailers and law enforcement has always been, ‘how do we help them focus on the highest harm offenders inflicting the most damage?’. 

How crime intelligence tech surfaces high-harm offenders

Over the past 12 years, Auror has helped to build a global network where what were once considered one-off incidents and isolated offenses, are now linked to the same people or groups, responsible for multiple events across different stores, retailers, cities, counties, states, and even countries.

By harnessing the power of the information retailers have always gathered, they can make better decisions about target hardening, but are also better able to help police to do their jobs more efficiently and effectively. 

That information has allowed the sector understand the contours of an issue that was once hidden through underreporting:

• Globally, 1 in 10 events involve some form of violence or aggression.
• 10% of offenders are responsible for more than 60% of the crime and harm.
• Repeat offenders are more than 4X more likely to be violent.

To address the scale challenge of retail crime, technology-based solutions that equip teams with the right intelligence to respond and act are vital.

That’s why we’ve made the core mission of our work, in partnership with many others, to reduce violent retail crime by 50% in the next five years. We call it 50 in 5.

We are going to be throwing everything at this challenge - new partnerships, new technology and advocacy. It’s about using technology effectively to deal with the volume challenge and importantly it's about helping retailers to more directly prevent harm as well as focusing on holding repeat offenders to account. 

‍So, what role does recognition tech play in this next phase of retail safety?

To reduce violence at scale and keep people safer, it's clear we now need more preventative measures that give frontline workers advance warning of known violent or prolific offenders, empowering them to make informed decisions about team and customer safety.

There is no one answer here, but we believe recognition technologies are a key part of the solution, provided they are deployed safely, thoughtfully and transparently.

Auror’s co-founder and CEO Phil Thomson recently shared that retail leaders around the world tell him why exploring facial recognition in particular is important, quoting one retail CEO as saying “I have an obligation to my team…to keep them safe. If technology is going to help us with that, why wouldn’t I look at it?”

Facial recognition technology (FRT) in the past three years in particular, has rapidly advanced into an effective, accurate and reliable tool. 

Training data sets have become way broader and more diverse, making a considerable leap in accuracy and anti-bias. The caveat is that image quality remains the most important input. But, if images are of good quality, then world-leading FRT is proven to deliver a true match rate of over 99.8% in challenging environments with poor lighting and off-angles, as demonstrated by independent testing organizations such as National Institute of Standards and Technology.

At Auror, we’re exploring how this recognition technology can be safeguarded to ensure it’s used securely, responsibly, and that it is effective and proportionate to the challenge.

Simply put, the violence has become much worse, the technology has become much better and the obligation to keep people safe has become much clearer.

That doesn’t mean Auror is building FRT - there are many world-class providers out there. It means giving retailers, if they choose, the right safeguards, processes and security, so they can confidently and ethically manage the technology with transparency and integrity. 

It means ensuring these tools can only be used for crime prevention and safety purposes, protecting privacy, removing any ability to track, categorize or profile groups of individuals. 

Importantly, it means both removing human bias, while ensuring a human is always in the loop. 

The value of FRT has already been demonstrated across many countries. It is well understood and used in the UK. Last year, a New Zealand supermarket, in only a limited area across their stores, prevented serious crime events such as assault and abuse by 130 incidents.

Recognition technology for risk detection and prevention is also already used by many retailers for vehicle license plate detection. When a known vehicle associated with a dangerous or high-harm offender pulls into a retailer’s parking lot, the relevant frontline retail worker receives an alert, giving them crucial minutes to determine how to respond.

These technologies won't be right for every retailer. Every organization needs to make a decision based on the challenges they face and the communities they serve. What’s clear is that retail leaders are looking at every possible means to keep their people safe and that means preventing violence and organized crime in their stores, as well as reporting it effectively when it occurs. 

Retailers are well aware that high-harm offenders operate across a wide geographic area, not respecting provincial borders or store brands and they are obliged to use the information and tools they have in the most effective ways possible, across their sites, to ensure their teams go home safe and well at the end of their shifts.

To learn more about Auror’s 50 in 5 movement, visit our webpage, or get in touch.