A London based visual Artificial Intelligence (AI) company, Cortexia, is developing a way to analyze suspicious public behaviour, with the expectations of a safer and secure environment. The technology is introduced at a time where the development of surveillance technology is promising to enhance and add extra layers of intelligence to the surveillance sector.

The company has previously developed technology which recognizes brands and clothing is now hoping to make a mark in the field of public sector surveillance. The surveillance technology which is still in the developing stages, couples AI with advanced digital vision. The AI system then processes footage captured by cameras via intelligent monitoring and automatically looks for pre-programmed red flags.

The firm’s chief solution officer, Alstair Harvery, calls the AI technology ‘action recognition’, due to its effective motion detecting capabilities. For example, the AI is trained to pick up movement such as running but can categorise between running in general and running in panic, leading to near accurate predictions. He says, “when you can recognise actions, you can then put them against scenarios, depending on the environment. That scenario and those actions can move towards predictive analytics.”

The AI technology is highly trained and reaches the same standards of humans who work with security cameras, with the bonus of being active 24/7.

Another interesting fact about the software is that it doesn’t recognise facial features and identities which may lead to bias, the concentration is solely on monitoring actions and spaces. Cortexia is currently putting the AI software on trial in the construction sector in order to remind workers to wear their helmets and protective clothing. The camera recognises items like oxygen tanks, and hard hats in real time, which in turn alerts and informs the business about those who haven’t got the right equipment on.