ACCORDING to a Queensland Police report obtained by the ABC under right to information the use of face recognition at the Commonwealth Games in 2018 “…lacked the data to operate effectively”.
The findings of the QPS report are of real interest to security managers and law enforcement personnel – as well as to suppliers. They highlight the organic nature of analytics and suggest planning and process are the key to meeting operational goals when defending high profile events.
Vital for analytics to work effectively is data – the more, the better. For the Commonwealth Games, police records were included but images from Queensland’s Department of Transport and Main Roads were left out. Effective use of the system was about more than software or integrating systems – it extended to legislation.
“Difficulties were experienced in data ingestion into one of the systems with the testing and availability not available until the week Operation Sentinel (the Games security operation) commenced,” the QPS report found.
“…not having the legislation passed, both Commonwealth and state, in time for the Commonwealth Games, reduced the database from an anticipated 46 million images to approximately 8 million.”
The facial recognition system used in Queensland during the Gold Coast Commonwealth Games also had so few specific targets that it ended up being used for general policing from a mid-point of the operation. Only 5 identities out of the 268 requested could be established.
“Given the limited requests from within the Games, opportunity to conduct inquiries for the general policing environment was provided to enable better testing of the processes and capabilities,” the report said.
Face recognition in public surveillance applications is still in its early stages of development and the capacity to handle the processing demands of simultaneous recognition across hundreds of cameras covering crowded places is not easy to achieve. The challenges faced by QPS are likely to provide valuable direction for law enforcement agencies looking to leverage face recognition solutions in the future.
#securityelectronicsandnetworks.com
