Milestone Systems: Reshaping video technology with AI

Kevin Müller, Milestone Systems, reflects on the key themes discussed during the company’s XPerience Day UK in September.

An expert panel at the recent Milestone Systems XPerience Day (MXD) UK at The Belfry delved into the transformative impact of AI and video analytics on the security industry.

Moderated by James Thorpe, Editor of ISJ, panellists drawn from across the industry discussed AI’s role in enhancing business operations.

They considered the use of AI and analytics to predict incidents and improve the customer experience in sectors as diverse as retail and logistics.

In a wide-ranging discussion, the group also highlighted the importance of ethical AI and responsible technology and the need to keep human beings at the heart of the process as AI-driven applications are developed.

In this Head to Head Exclusive, Kevin Müller – Panellist, Event Host and Sales Director, Northern Europe at Milestone Systems – reflects on the importance of this discussion.

How are AI and computer vision technologies transforming the security industry?

AI and computer vision techniques are already impacting the security industry and present a great opportunity for the sector in the future.

With millions of networked video cameras already installed, incorporating AI into their video surveillance systems allows organisations to not only better identify and manage risk, but also gain deep insights that lead to greater operational efficiencies.

This presents a massive shift from only a few years ago when cameras simply captured video footage. Now, video cameras are the single most deployed IoT sensor around the world.

They add business intelligence to key processes and allow users to not only manage risk after an event, but even predict the events before they happen.

This has been driven by a massive acceleration in compute power and panellists illustrated this through the arrival of cost-effective yet increasingly powerful GPUs.

These allow self-driving cars to gain the contextual understanding of space and time necessary that in turn makes autonomous driving possible.

This means equipping vehicles to not only identify when something happens but understand why it has happened and consider how to act on the event.

The audience heard how such greater contextual understanding is becoming more prevalent in video technology. It will be truly transformational.

Panellists expressed their belief that integrating this powerful AI with IoT devices and sensors such as cameras will allow organisations to take more proactive security measures.

The audience were referred to the retail sector, in particular, where AI-driven analytics are enhancing security, improving the customer experience and providing greater operational insights – blurring the lines between security and business intelligence.

Of course, another driver of AI and analytics is an open platform such as Milestone’s XProtect, which helps developers create applications that leverage the latest developments in AI.

It also helps them get solutions to market to meet specific customer needs, confident that an ecosystem of compatible sensors ensures interoperability, cybersecurity and reliability.

At Milestone’s UK MXD 2024 event, many of the technology partners who work within Milestone’s open platform were demonstrating such cutting-edge AI applications.

How do we ensure that AI technologies are developed and applied responsibly?

Everyone on the panel at UK MXD 2024 agreed that as AI continues to be developed at lightning speed, its ethical use – and the responsible use of wider technology around it – is essential.

The risks of unchecked AI deployment, the potential for bias and inaccuracy, for example, are real. Businesses therefore have a duty to ensure AI applications are ethical – and this means that human involvement is essential in AI decision-making.

Even when AI-driven video technology detects an incident, having a human ‘in the loop’ to ensure an appropriate response is vital.

Alongside organisations’ commitments, regulations will help drive the ethical and responsible use of AI, of course.

Milestone Systems has led by example here, and earlier in 2024 adopted and implemented the G7 Code of Conduct for advanced artificial intelligence (AI) systems, becoming one of the first companies to do so.

As Milestone’s CEO, Thomas Jensen, noted at the time: “We need rules to ensure AI is being developed to serve humanity. But companies should not wait for regulation.

“They must take their own steps to identify and resolve the weaknesses and pitfalls of the AI they develop.”

How is synthetic data being used to mitigate bias and enhance the accuracy of AI models in security applications?

During the discussion, myself and the other panellists addressed the potential for bias in AI models and the increasing expense and rarity of large and suitable models on which to train AI.

The audience heard how synthetic data can be used to reduce bias and improve the accuracy of AI models.

I shared a real-world example of the use of synthetic data from work at Aalborg University in Denmark, where researchers used synthetic data to train AI to recognise a person falling from a harbourside into the water.

Using a synthetic dataset to supplement thermal images and a dummy, researchers were able to train the AI-enabled video software to accurately identify falls, so the system could alert rescue services in the event of a fall.

These types of exciting applications for AI are not only great leaps forward in technology.

They present significant opportunities for those that have traditionally relied on video technology to provide passive security, but who now see the massive potential of real-time operational insights and predictive risk management to their organisations.

What was discussed when it comes to the integration of AI?

At the end of the session, panellists urged those present to grasp AI as a means of improving their organisation’s operational efficiency, enhancing security and creating new commercial opportunities.

To help guide them on what might appear at first to be a daunting journey, here are some simple recommendations. Think of them as action items that will help your organisation reshape its use of video with AI:

• Seek out opportunities to use AI-enabled video technology to gain business intelligence and insights
• Educate customers on the benefits of embracing AI technology and becoming an expert in the field
• Collaborate with technology partners that understand AI to make solutions that are easy to implement and deploy
• Ensure the ethical use of AI is ‘baked in’ to your activity and that this is driven by the regulatory environment, but also business principles

Milestone’s UK XPerience Day 2024 Panel was hosted by the Editor of International Security Journal, James Thorpe.

To find out more information about upcoming EMEA Milestone XPerience Days, visit: www.milestonesys.com or contact Leigh Domingo-Hughes [email protected]