Vincent Hotel, Southport is a luxury.
The six floor boutique hotel, opened earlier this year, has 60 guest studios, residences and penthouses (they don’t call them bedrooms – a penthouse during this year’s British Open at nearby Royal Birkdale cost £15k). Golf simulator on the top floor, personal trainer, spa, internet access, wall mount HD – LCD TVs, work-stations, iPod connectors, an MP3 collection with thousands of titles: you name it. Here, as in any other hotel or retail generally, there has to be a balance between discreet and effective security. A hotel does not want to pat down everyone that comes through the door; equally, a hotel does not want security so loose that thieves take advantage of the welcoming environment.
Seeking to keep clients and their valuables secure, hotel management contacted Human Recognition Systems (HRS). Typically Liverpool-based HRS has installed biometrics into airports; construction sites; prisons; and dispensing pharmacies to recognise patients. At The Vincent, HRS fitted facial recognition to CCTV. Lee Hannis, HRS opportunities manager, told Professional Security that the system will trigger alerts, according to the rules set by the user. He stressed how people, policies and processes have to work with the technology – put another way, if staff are not trained in use of the solution, if processes are not in line with the way the business is run, the technology will not ‘work’. For instance, typically an alert will go to a central monitoring station. It’s for the user to set how many minutes or seconds pass before the alert is then escalated to a supervisor or the next nominated receiver, if not answered.
Uses may be strictly security – rather than hotels like shops briefing staff with a watch-list of suspects, and relying on staff to retain a memory of the professional thieves’ faces, the facial recognition software checks the face on entry. Or, a CCTV camera can cover a painting or other high-value item, and software can send an alert if the object is absent. Again, this has to tie in with the hotel’s operations; it may be routine that a painting is moved for dusting.
Generally speaking, Lee separated biometrics from what he called ‘human behaviour analysis’, which he said seemed a logical extension from a biometric solutions such as iris or fingerprint recognition. The much-described ‘intelligent CCTV’ can view a scene and if there is anything out of the norm – or pixels static for a set time – an alert is sent. As Lee said, that does not necessarily classify the object. Where the human behaviour analysis comes in: recognising that the object is a car, or a loitering person, based on human geometry, of a head, legs and arms, rather than say a trolley in a corridor. "Once we have recognised that there is a human in the picture, we can then extract the face," and bring in the biometric recognition – in this case, identify the face.
Lee Hannis agreed that a user could find non-security, general operations uses. A hotel usually does not have a member of staff dedicated to CCTV; footage is trawled through only after a theft or stock loss. The behavioural patterns set by a user – and alerts if there is anything out of the ordinary – can include stock rooms, if they are entered at an odd time of night. Rather than a stock loss only found at the end of the month – costing staff time in investigating besides – a manager can investigate the alert at once.
