Developments in artificial intelligence (AI) was the decisive factor that made this possible as algorithms now help cameras understand what they capture.
Long before COIVD-19 forced governments to enforce restrictions on companies and people, the video surveillance industry had begun exploring ways to decrease workforce needs. Developments in artificial intelligence (AI) was the decisive factor that made this possible as algorithms now help cameras understand what they capture.
But the use of AI-based cameras is still not as popular among end customers as they should be. The
benefits are high, but customers are yet to adopt the technology. While this could be attributed to concerns on ROI for some, some systems integrators argue that the problem is that there is a gap between vendors’ presentation of AI-based systems and the customer’s understanding of the same.
For instance, an advert of an AI-enabled camera would include a description of its features. But the customer needs to know who these features make a difference to their bottom lines. This is where SIs also need to play a crucial role. Here are some examples that some systems integrators have pointed out where customers
benefited from AI.
Meeting needs of housing societies
Large residential societies need security cameras to protect people and assets. But what if the cameras could do more? This was something Patrick Lim, Group Head of Strategy at Ademco Security, pointed out. For instance, residential societies have complex piping systems, and these need to be monitored to ensure that there are no leaks or damages. Traditionally, an employee would have to follow this periodically, making it a tedious, inefficient, and expensive system.
AI-enabling algorithms can detect issues such as leaks and equipment failure without human intervention. They can alert the people concerned as soon as a problem is detected, enabling a quick response.
Monitoring product count
When an Indian company manufacturing pipes wanted a system that could count the number of products it loaded on a truck, their SI, Securizen Systems, had a solution.
Sandeep Patil, Founder of the Securizen Systems, explained that generally, when there is an order for products at such a manufacturing plant, it includes different kinds of products, like a 2-inch pipe, 5-inch pipe, 10-inch pipe, etc. While loading these products, for better space management, the company would place a 5-inch pipe inside a 10-inch pipe, a 2-inch pipe inside a 5-inch pipe, and so on.
The customer wanted then was to know the number of products loaded on the trucks through video surveillance cameras. With the help of AI, Patil’s team could identify how many types of pipes were being loaded on a truck. This would get automatically reconfirmed with the order that was received.
Detecting danger from a position
Sites like airports and critical infrastructure need to continually watch out for intruders while maintaining a welcoming environment for passengers and staff. But given the large number of people who are at a place like an airport at any given time, human monitoring of every person becomes difficult.
For instance, an intruder with a concealed weapon can reach as far as the check-in counters in many airports without being checked. If he draws a gun after arriving there, security officials would already be late in responding to it. This issue is not limited to airports. Even large companies or campuses need to maintain vigil against such intruders with weapons.
This was an issue that Verghese Thirumala, CEO of Malaysia-based SI Maxitulin, had to deal with for one of his customers. He was able to deploy an AI-based analytic solution that would detect a person when his body is in a position of holding a weapon. In other words, a person may enter a premise with a concealed weapon, but as soon as he draws it and positions himself to use it, the algorithm would detect it and alert the security.