Big 3 Themes From ISC West 2024.
Big 3 Themes From ISC West – ISC West’s big 3 themes were artificial intelligence, SaaS, and robots and drones, writes Optic Security Group product innovation manager, Vlado Damjanovski.
The International Security exhibition and Conference – otherwise known as ISC West – is one of the largest events in the world dedicated to the security industry. This industry covers CCTV, Access Control, Alarm Systems, gates and other security related products, the event is traditionally held in Las Vegas every April.
This year, ISC West attracted over 700 exhibitors from all over the world, and, according to many veteran exhibitors, it was the biggest year in the show’s history. The latest and greatest hardware and software products are exhibited first at the ISC West, making it the place where the latest trends can easily be seen.
The companies dominating this year’s show in terms of floor space and advertising presence included Hanwha Vision, Motorola Solutions, Axis Communications, Bosch Security, Johnson Controls and Dormakaba. Smaller companies with smaller stands, typically around the perimeter of the main floor, offered some new and innovative products, and they grabbed my attention too.
It was impossible to visit all stands due to the sheer size of the exhibition. To make things a little easier, the organisers of the show have developed an app for quick locating of exhibitors, which presents a walking map based on search criteria to help attendees find their way.
Optic Security Group sent me across to Las Vegas to gain first-hand perspectives on what was on display, to chat with the experts on the ground, and to identify the innovations that attendees were talking about. Below are my big 3 takeaways.
Artificial Intelligence (AI)
Without any doubt, the most often heard acronym at the show was AI. Throughout the main exhibition floor, AI-capable products – cameras, hardware, as well as VMS (video management systems) software – were the most common theme and attraction.
There were many existing and proven VMSs, but also some new and potentially very attractive newcomers, all attempting to show how they have mastered AI video analytics. VMS software typically embeds AI based algorithms for a variety of applications, but solutions also included those in which the AI analytics is embedded in the cameras themselves (edge-analytics).
AI algorithms have certainly matured, but the one limitation that has become obvious is that most companies manufacturing cameras or implementing software solutions for AI analysis of video streams use a standard set of “Object Classification”. This means that cameras or VMSs have the ‘intelligence’ to immediately recognise pre-defined (or pre-learned) objects, such as humans, vehicles, trucks, busses, and bicycles instantly, but not much more. There remains the need for AI to be able to recognise other non-standard objects, but this is easier said than done.
Some companies have now started offering AI solutions in which customers’ objects of interest can be ‘inferred’ (taught) in the AI analytics (such as Hanwha’s Flex Ai). This is not a straight-forward method, as the so-called ‘inference’ requires hundreds or thousands of examples that need to be entered manually for the algorithm to learn. In addition, the computer processing power required for this learning is substantial and – in the main – the entering can only be done by the AI manufacturer as opposed to the user.
It is encouraging, however, that some companies have started offering the ability for users to ‘add new objects’ in the object classification table.
It was also interesting to see that some AI manufacturers have started offering behavioural analytics
(Ambient.ai). Once recognised as a human object by the AI, the next step is recognising an individual’s posture and – combined with the speed of limb movement of two humans in close contact – the recognition of human fighting. This was demonstrated to good effect.
Security as a Service (SaaS)
Quite a few companies are offering security as a service (SaaS). Basically, instead of a customer investing in recording hardware and analytics, SaaS service providers offer subscription-based storage and analytics solutions. In short, users can subscribe to a plan that uploads their camera footage streams to the cloud for both recording and AI analysis by a SaaS provider’s powerful computers.
Under such models, customers pay a subscription for their selected AI functionalities and the length of recording. The advantage to the customer is that they don’t need to worry about the technicalities of the solution, how much storage they need to buy, or how powerful the computer server(s) need to be. This is all taken care of by the SaaS, for a price.
In addition to the likes of Verkada and AWS, SaaS products are offered by most of the major vendors, such as Motorola (having recently joined forces with Google Cloud), Genetec SaaS, Hanwha Cloud, etc.
It appears to me that SaaS would be an attractive security solution for a small household user, and this seems to be increasingly the case in the US, UK and Australia.
A critical requirement for SaaS-based video surveillance is sufficient bandwidth for uploading video streams of all cameras at a site (such as a domestic residence). Each camera stream takes at least 1Mb/s – 2Mb/s of bandwidth of internet connection, so for a 4-camera system for example, at least 8Mb/s is required for continuous uploading streaming.
With household internet connections the uploading speed is always slower than the download, and depending what type of connection a household has, it may not be sufficient.
In most industrialised countries, this might not be a problem, but the streamed amount of video data, even if the bandwidth is sufficient, is paid for by the user in addition to the SaaS subscription. The SaaS video streaming for recording and AI analysis is required 24/7, so data needs to be constantly up streamed.
Robots and Drones
There was a notable presence at ISC West of companies that make autonomous and remotely controlled robots for security purposes. Deployments might include checking areas in and around shopping centres after hours, patrolling with robot cameras, carrying illumination, speakers with pre-recorded warning messages and 2-way communication with the operator.
Robots can also be used in dangerous situations where there are suspected explosives, or where a human security presence might otherwise be at risk.
I saw 4 wheeled police/security robots (Code blue and Argus), but also dog-like robots (Go2) walking or jumping on 4 legs, able to move across various terrains, including up and down stairs. These robots are powered by internal batteries that are sufficient for 4-8 hours of operation.
A manufacturer with noticeable drone advancement (code blue) demonstrated self-parking and self-charging drone stations. Knowing that all drones are limited to flying for no more than 30-50mins, they need quick and automated recharging. The ‘drone station’ solution features covers that open for the drone to fly in/out.
Before the drone’s battery is depleted it automatically returns to its nest and starts recharging. The drone station closes to provide protection from rain, snow, or unwanted interference.
Other interesting solutions relating to drones included drone detection software that utilises a PTZ camera for visual identification (Drone Object Classification) of unwanted drones in an area. There are numerous applications for such a device, such as airports and prisons, where drones are often used to smuggle drugs, mobile phones or weapons.
In conclusion, being at ISC West offered an experience that is impossible to capture in a single article. Nevertheless, I hope you find my few insights useful. My trip to Las Vegas for the big event was made possible by Optic Security Group in my role as Optic’s product innovation manager. If you haven’t had the opportunity to attend, I recommend that you do.
You can learn more about Optic Security Group here or read more SEN news here.
“Big 3 Themes From ISC West 2024.”
