Mike Poe, Director of Product Management, 3xLOGIC, explores how organizations can take advantage of the data captured by video analytics.

The possibilities with video analytics

In a world saturated with data, collecting it is no longer enough for a strategic advantage.

Instead, the best security leaders are leveraging their surveillance systems to gather, consolidate, analyze and respond to events and threats using advanced video analytics.

As this piece will explore, the potential uses for your surveillance system extend far beyond traditional security, providing greater peace of mind, operational efficiencies and value across your organization.

Surfing the data wave

Data has become the key driver of modern businesses, with large amounts generated daily.

By 2025, the World Economic Forum estimates that a staggering 463 exabytes of data will be produced daily.

To put that into perspective, every word ever spoken by human beings, in languages current and old, is said to take up only five exabytes, according to TechTarget.

With the rise of the internet of things (IoT), including connected, intelligent video cameras, the amount of data produced will rise exponentially.

Transforming data into insights

It isn’t just about the quantity of data available at your fingertips but the quality – and the computing power to analyze it.

Data, in its raw form, is unusable by the average security team.

It needs to be transformed into actionable insights for the organization, through data mining, analysis, data visualization and the right data infrastructure.

This is business intelligence and it’s becoming vital in the security sector.

Technology making this possible

Advances in AI in security systems mean it’s easier than ever to have an intelligent system scanning for unusual behavior and potential threats.

Some of these analytics can even be carried out ‘at the edge’ on cameras themselves, removing the need for expensive servers and related computing infrastructure, creating more real-time alerts and actions as cameras can respond immediately to objects and events detected.

The level of detail captured in modern-day cameras enables more advanced analytics.

Clear footage is needed for AI to analyze effectively, so cameras are designed to capture high-resolution images in a range of environments, indoor and outdoor, in temperatures up to 138^oF, in inclement weather and low lighting.

This also delivers clear images for post-event investigations.

Detailed insights

Improvements in cameras and video management systems (VMS) are also unlocking more detailed insights from footage.

Cameras can now use object and people detection, vehicle identification and attributes like color and model to accurately detect and track vehicles – even in complex, busy environments like a train station concourse or shopping mall.

Using vast amounts of representative data, the cameras can continuously analyze movements and behavior, while using predictive analytics to track each object continuously and automatically pass onto another camera in the system if an object moves out of range of the original device.

When an object is detected or an event occurs, the VMS can alert human operators to respond as needed.

Insights can be shared with staff to inform their approach or operators can follow an object (i.e., a person’s) movements across a space to understand more about what’s unfolding.

The role of deep learning

Deep learning is an important piece of the puzzle here, as it reduces the number of false positives that can frustrate and fatigue security teams.

For example, traditional systems may have alerted operators to moving foliage or illumination changes.

With deep learning, the tracking system can identify these movements as such and ignore them because they don’t represent objects that demand investigation.

This makes the system a lot more accurate and helps to ensure operators respond to relevant events instead of ignoring alerts because they’ve had too many false alarms.

This level of filtering is also highly useful for complex situations and behaviors.

Operators can gain insights into a person’s dwell time within a specific area, or traffic direction moving across a logistics yard.

Tailgating can be detected, along with abandoned and removed objects.

Advanced video analytics can also provide accurate information about an object’s entry and exit, appearance and disappearance.

These insights can support access control and perimeter protection, with accurate intruder alerts, the ability to track visitor numbers and movements, and alert to unexpected activity out of hours.

Expanding beyond security

However, security systems are no longer just about surveillance.

When implemented and used effectively, your video surveillance can be a strategic asset to every part of your organization from sales and marketing to operations.

People counting analytics, for instance, can influence many aspects of a retail environment, from where high-value stock is displayed to how many staff are on the schedule.

Alerts can be issued when queues are detected, allowing staff to proactively open another check out.

Over time, management can analyze how footfall changes on different days or weeks and plan schedules accordingly.

In an office setting, understanding how a building is being used by employees throughout the week can influence the heating, ventilation, lighting and maintenance schedules – driving greater energy and cost efficiencies.

In a warehouse, video can alert to potential hazards, such as a forklift moving in an area close to human workers.

If a human worker is detected as not wearing appropriate PPE (personal protective equipment) like a high-vis jacket, then the system can alert operators to remind that worker.

Dangerous behaviors, such as jumping under roller shutters to gain access or exit, can also be flagged.

In the near future, video analytics will advance to a state where it can proactively warn operators of equipment failure, preventing unnecessary downtime.

Of course, for this level of intelligence to be unlocked for your organization, there are some foundational elements you need to be aware of.

Putting the right groundwork in place

Training your security team to work alongside advanced video analytics and trust its insights is vital.

Some may be less familiar with the technologies being implemented today and may fear or distrust it, particularly when it comes to AI.

Training can show operators the benefits of advanced video analytics and how such tools will augment, not replace their jobs.

It can also provide the skills needed to critically assess the alerts coming in from cameras and to set up new rules and conditions to expand the video system’s capabilities further.

Additionally, early collaboration with your IT department will help your implementation run more smoothly.

By involving them from the outset, you can streamline deployment and ensure seamless integration with existing infrastructure, setting the stage for long-term success.

The technology being used needs to feel intuitive and easy to understand, to reduce training needs and also encourage operators to use it more.

A complex tool won’t inspire confidence or motivate people to explore its data.

That’s why clear data visualization is a key part of good business intelligence.

As computer scientist Professor William Cleveland explained in his book Visualizing Data: “No matter how clever the choice of the information, and no matter how technologically impressive the encoding, a visualization fails if the decoding fails.”

You can gather all of the video data available to you, from across every part of your business, but it won’t create value or impact if your people cannot understand it.

Forging a data-driven path forward

The market leaders of 2030 will build their advantage on a foundation of data and business intelligence.

They will decode data, unlocking insights that improve not just safety and security but operations, strategic planning and more.

Business intelligence that uses your video data offers unparalleled visibility into the everyday happenings across your organization.

That information shouldn’t solely be kept within security but offered to your counterparts in other teams so that everyone benefits from the rich insights that modern cameras and a VMS create.

This article was originally published in the June edition of Security Journal Americas. To read your FREE digital edition, click here.