AI slashing security response rates and mitigating risks
What is the impact of AI in the security technology market, in particular on CCTV equipment?
For many years now Artificial Intelligence (AI) has been a buzzword. However, if the most frequently heard phrase during the pandemic was ‘you’re on mute’, then ‘artificial intelligence’ is surely making a comeback, and with a bang!
The technology available to the market today is becoming increasingly more intelligent, and following the impacts of the pandemic and the challenges that came with it, we have seen a significant shift in the use and requirements of AI technology within our security technology. The top areas of AI that are commonly explored and discussed include Machine Learning, Deep Learning, Natural Language Processing, and Robotics.
AI is sophisticated and far-reaching. What is certain is that it plays, and will continue to play, an increasingly significant role in our daily lives, influencing various aspects and making numerous tasks more efficient, personalised, and accessible.
I recently attended a very informative ‘Counter-Terrorism: Cyber, Safe Cities and Advanced Technical Event’, and the following statistic struck me: ‘Mark Zuckerberg took 10 months to reach 1 million users in 2004, whereas ChatGPT achieved the same milestone in just 5 days’. Take a moment to consider the sheer explosion in the volume of users turning to AI.
AI and CCTV
Closed-circuit television (CCTV) surveillance systems produce large amounts of video footage, most of which never gets reviewed. As a result, security incidents are missed, and suspicious behaviour isn’t caught in time to prevent incidents. AI analytics can help address these issues.
AI-based analytics use algorithms to go through both live and recorded video content and generate descriptions (metadata) of what is happening or has happened at a particular location. Objects such as vehicles and humans can be detected and then monitored to find significant events you want the system to react to automatically. These events can be anything from a person detected in a prohibited access area to a vehicle approaching a gate. This filtered data enables security personnel to efficiently access and react to situations.
An abundance of information is readily available – the real challenge lies in effectively utilising this information to enhance and support our businesses to mitigate risks.
CCTV Post-Security Incident
Historically, CCTV has been used as part of a post-incident review process to analyse footage captured or construct a timeline of events. This is a laborious task, with many hours spent trying to build a trail of information and reliance upon third-party information. Even within a live control room environment, it is increasingly difficult to identify incidents within busy locations. The workforce required to manage this environment can often outweigh the benefits.
AI enables the ability to search for specific items such as clothing, colours, and search functionalities. Hours, if not weeks, of footage are now accessible in minutes, at the press of a button.
The ability to share important information not only swiftly but also securely, with audit trails and encryptions, is also a benefit AI has brought to the post-incident review process.
I have worked in the Electronic Security Industry for 21 years now, and one of the most important phrases that has always resonated with me is ‘to create a proactive approach to a reactive environment’. This has only been accelerated since joining ICTS, as the impact and support we can offer with technology and AI to frontline staff can be the difference in a critical situation.
CCTV Pre-Security Incident
In recent years there has been a shift to the usage of specific AI software for the monitoring of loitering, certain behaviour, people counting, heat mapping, and many more data-gathering uses. AI has assisted in the creation of early alerts for hostile situations and thus enhancing overall security.
Data is key
Collation of data, and specifically targeted data, that can assist in a proactive approach is critical in any risk analysis. Utilisation of software-driven intelligence and accurate creation of rules can be the difference between a successful deployment and early intervention.
Analytics based on deep learning provide more accurate results and eliminate the need to continuously monitor video footage. Whilst deep learning-based analytics can help security personnel ‘see’ and limit human error and bias, it cannot entirely replace operator experience and decision-making skills.
Expenses with the traditional ‘remove and replace’ approach
The added benefit of utilising AI in today’s technology-filled market is that in many scenarios software can be installed on existing hardware without the need to replace hardware.
In many situations, there is a misconception that introducing new technology comes with a large investment. There is always a cost to any new strategy or solution, but in my opinion and experience, it is often very easy to provide accurate and proven examples of an extremely quick ROI (return on investment). The ‘spend to save’ scheme is widely acknowledged within the public sector, SMEs and even the largest FTSE 100 or Global 500 organisations.
Human Intervention
It is important to highlight that no matter how intelligent we get, AI will always involve some form of human intervention. We are often reminded of this by the tech companies.
The impact of AI on security technology is significant. AI-enhanced CCTV provides a valuable example of how the advances in technology can make our environments more secure and our security operations more efficient. Return on Investment is rapid, especially as some AI software can be installed on existing hardware equipment. However, whilst AI has become an integral tool in security technology, it cannot entirely replace a security professional’s experience and decision-making skills.
Curtis Wright
Director
Electronic Security and Monitoring Services, ASM – the technology subsidiary of ICTS UK & Ireland.
