How intelligence evolves the future of security

The future of intelligence


The future of security intelligence

Looking back, the world we live in has been quite static, slowly changing day by day. People have been shaped by evolution to cope well with this environment, developing intuitions to deal with our everyday reality.

But times are changing and with the introduction of digitization and an explosion of data, our environment is changing so rapidly that people can no longer rely only on intuition to navigate through reality – or more importantly, to address the security challenges of today. This is where data and intelligence comes into the picture.

The future of security intelligence is about generating, capturing and analyzing data so that we can predict and prevent risks and future incidents. Combining the power of machines and algorithms with the human mind will allow us to create a safer and more secure environment for everyone.

Data driven risk analysis

Using intelligence to foresee risks

In the good old days, risk analysis was static, performed once or twice a year on a global level. However, changes in the local environment happen all the time, soon making the analysis obsolete. To make our risk analysis more valuable, we are adding data and analytics to the mix, to be more responsive and detailed in our analysis.

Our dynamic risk analysis collects data from both internal and external data sources. This includes incidents, various sensor data, crime statistics and weather data. As an example, crime statistics, weather data and visual monitoring can be combined to get information about if an intrusion alarm was triggered due to a storm or if there was an actual intrusion. This helps us to reduce false alarms and to and allocate resources to situations where they are needed the most.

With data based risk analysis we can be much more detailed and granular in our conclusions, pinpointing risks all the way down to certain geographical areas, type of site, and even type of crime.

This makes us radically better prepared to handle a likely future. We are also looking into adopting machine learning in our analysis, so that we can create efficient, self-learning algorithms that not only predicts the future but becomes better the more data that is fed to them.

With analytics and data in the background, we augment our security officers’ capabilities in the field, so they can focus on what humans do best, at the right place, at the right time.

Augmented security

Augmenting security with augmented reality

At Securitas, we are currently exploring, together with Microsoft and Combitech, how augmented reality technology can interact with humans to improve our security delivery. We are working on a prototype solution based on Microsoft Hololens.

It is a set of AR goggles, that displays a data powered layer of reality to reinforce people’s already sharp minds, effectively linking the digital world with the physical. Let’s explore what we could do using the solution.

Imagine a threatening situation in a shopping mall – a brawl or a robbery. Time is critical for the security officer to reach the situation. The hololens solution guides our security officers through the mall, by calculating and then displaying the fastest route in the goggles. This saves valuable response time, time that can be the difference between disaster and success.

security officer with VR glasses

What if a security officer could keep track of things disappearing or moving? With our solution, security officers can detect changes in the physical environment that would be virtually impossible for humans alone. A fire extinguisher that has been moved, a door that has been opened or a car that mysteriously has disappeared, the solution can alert security officers of these slight changes and allow them to act.

The goal is always to create a safer environment and data can obviously play a big role in this. But we will never be slaves to data, rather we see it as a way to empower our people at the frontline. People are always at the center because when it comes down acting, resolving threats, we are superior to machines.

Intelligent retail

The intelligent retail customer journey

In a retail context, data and intelligence can play a dual role. On the one hand, we can use it to create a safer store environment and on the other hand we can leverage data to improve sales and the customer experience.

Let’s take a look at a data powered customer journey through a mall. Imagine a person, let’s call her Lisa, visiting a shopping mall to purchase a new sweater. Lisa has a loyalty card connected to a specific mall and choses to go there, instead of the local mall.

car seen from above

When Lisa arrives at the parking lot, the mall identifies her through license plate recognition and opens the gates to let her in.

Lisa has downloaded the shopping mall’s app, which also has registered her arrival and a personalized offer is pushed to her phone. Lisa goes to the store that pushed the offer, guide by the app’s map, purchases a sweater, pays for it and leaves the store. Her actions is monitored and used to even further personalize her experience for future visits.

While Lisa walks through the shopping mall back towards the parking lot, she sees a large crowd and realizes that a celebrity has paid a visit to the mall. She also notices that there are a lot more security officers than previously. Securitas has been able to extract intelligence from social media data, and has been aware of this situation long before the celebrity even arrived and could arrange for more security officers to secure the situation.

As Lisa drives out of the garage, the license plate recognition solution tells the shopping mall exactly how long Lisa has stayed and if she needs to pay for her parking or not. All the data that has been gathered during her visit will help the shopping mall further increase Lisas shopping experience in the future.

Woman carrying shopping bags and a cell phone

At the same time, the different sets of data that improve her experience also create improved safety.

For example, if a car has been in the parking lot for several hours, or been switching parking spots several times, this behavior can be detected so that a security officer can take a closer look.

All above use cases are built on analytics and rule based intelligence. We are now looking into adding machine learning powered intelligence on-top of this to create self learning algorithms so that the system continuously improves. Imagine a store during closing time, when cashiers often work alone and are vulnerable to assaults. In this context, sensors can track how long it usually takes to lock the store and when the alarm should be turned on. Using machine learning to track this, we can uncover behavioural patterns on individual level and understand if anything is out of the ordinary and if a security officer should look into the situation.