- Security TWENTY
- Women in Security Awards
Understanding AI should be your 2021 resolution, writes Harro Stokman, Founder and CEO of AI-based software firm Kepler Vision Technologies.
AI is no longer something just found in the cybersecurity space. The huge steps forward that the technology has made in the last 10 years have cemented it as an invaluable tool for enhancing physical monitoring and security systems. Increased availability has resulted in lower costs and greater user friendliness for the technology – you no longer have to be a computer expert to utilise it and benefit from the operational savings it can provide.
When we talk about AI in security we are usually referring to camera surveillance technology, specifically computer vision powered solutions like face recognition and person detection systems – all of which rely on trained AI to recognise different facial or body configurations. Here at Kepler Vision we apply this technology to look after the wellbeing of care home residents, using it to identify when a patient is experiencing difficulties and automatically alerting care staff to help. The same technology can be applied to identifying when someone is trying to break into a secure location or is engaging in suspicious activity.
While having a person monitoring a live camera feed can provide you with similar results, anyone with any experience monitoring a feed for hours at a time (with very little happening) will tell you just how mind numbing the process can be. Then consider the same person needing to monitor dozens of feeds, or hundreds. Human error becomes exponentially more likely the more a single person is expected to monitor. AI-powered video analysis is not a tool to replace these personnel, but rather supports these workers by flagging suspicious activity in real-time, making it easier to manage and respond to.
Motion detectors can also provide real-time notifications to homeowners or relevant staff when triggered, but their primary drawback is the volume of false alarms they generate. Computer Vision AI systems can identify any movement, but only alert monitors when the movement is actually relevant – separating routine behaviours or movement from abnormal ones. For instance our own “Kepler Night Nurse” system reduced the rate of false alarms for patients to a single false alarm per video stream per 186 days, massively reducing the workload of camera monitors. Responding to every instance of a triggered motion detector is extremely inefficient. It is far better to limit responses to situations that warrant action, and computer vision systems make this possible.
While this technology may be unnecessary for home security reasons, for enterprise level the benefits of an AI enhanced camera system are difficult to ignore compared to their more analogue counterparts. The sticking point is often justifying installing an entirely new system to replace an existing one, and all the expense that often involves. This is where computer vision’s more recent advancements come into play, as many offerings in this space can integrate seamlessly with existing camera systems, without requiring new expensive cameras or additional equipment. A cloud based solution can act as an overlay on existing video feeds, constantly “watching” the streams for the set of circumstances that would trigger an alert.
A well “trained” AI system can accurately distinguish between humans and vehicles from wildlife elements and other innocuous objects – while not flagging recognised individuals and vehicles as false threats. This added accuracy has the ability to make a huge difference to security workers’ day-to-day activities by giving these workers back crucial time spent responding to false alarms and physically monitoring camera feeds – presenting an opportunity for them to spend their time on other core competencies, and allowing them to react proactively rather than retrospectively.
As new tech continues to transform how we approach physical security, the challenge for decision makers will be in learning to evaluate the value and appropriateness of each tool for the job, and to filter through the industry hype. Even with the advantages that AI aided security systems can provide, it is important not to lose sight of the human component. Alerts still need to be evaluated and responded to by a living breathing person, just as before. But by combining, machine and human expertise, AI systems in physical security have the potential to be the most powerful instruments in your security manager’s arsenal.