Smart CCTVs: Third eye of secure cities

Advances in artificial intelligence are increasing the potential of intelligent, automated security cameras

Following recent terrorist incidents, Germany's Interior Minister announced in August that CCTV cameras at airports and train stations will be enhanced with facial recognition technology. Likewise, the New York Police Department has developed the Domain Awareness System that uses similar technology to track and monitor potential suspects.

Globalisation increases the exposure of cities to myriad transnational threats even as growing urbanisation is putting the strain on law enforcement by increasing the densities of population, property and critical infrastructure to be safeguarded in each precinct. These inherent challenges in protecting cities - population and economic centres that make attractive soft targets - necessitate the early warning and identification of threats. Smart CCTVs support this function as the third eye of cities by complementing the vigilance of police officers and the community.

CCTV surveillance of public spaces has been a routine security feature of urban environments since the 1990s but grew in ubiquity post-9/11. Premised on the concept of "defensible space", it is a physical expression of the community's ability to defend itself against perpetrators and over time grew in importance as the fifth utility alongside critical infrastructures: water, gas, electricity and telecommunications. Past incidents have demonstrated its utility in post-event investigations and disruption of further threats.

Advances in artificial intelligence (AI) are improving the accuracy of video analytics - facial, behavioural and object recognition - therefore increasing the potential of smart CCTVs to fully or partially automate the processing and analysis of voluminous data collected from a vast network of cameras and in long-term decision-making. Smart CCTVs are capable of round-the-clock, city-wide intelligent surveillance and are not subjected to human limitations.

Countries are increasingly embracing smart CCTVs as a quintessential feature of smart cities to meet evolving security needs, given the changes to the character of cities due to growing urbanisation. For example, the police camera project which deploys CCTV cameras extensively in residential towns is a key feature of Singapore's Smart Nation initiatives and enhanced counter-terrorism strategy.

Private security firms have also begun adopting the technology to re-engineer business processes by optimising security patrols with remote surveillance of their clients' properties.

Smart CCTVs will henceforth have a critical role in the coming years in securing cities as well as in homeland security. Their proliferation would expectedly raise privacy concerns, and their omnipresence could inadvertently create the illusion of gated communities and increase complacency in terms of personal security.

While smart technologies are expected to bring benefits to modern cities, they also introduce vulnerabilities. Interconnectivity by nature enlarges the potential attack surface of cities and reveals novel attack vectors for threat actors to exploit. The hacking of the police-operated CCTV system during the 2015 South-east Asia Games in Singapore demonstrated the plausibility and criminal intent to target law enforcement agencies.

In February, Hizbollah television station Al-Manar's claims that the militant group had hacked into CCTV cameras in Israel demonstrated a hostile intent to undermine CCTV systems as part of a larger information warfare to undermine the Israelis' sense of security.

Therefore, the spectrum of cyber attacks on a city's smart CCTVs could range from sheer criminality to compromising national security, given that cyberspace is the fifth dimension of warfare, and cities are the lifeblood of nations.

Cyber-security risk management should begin with assessments of the four aspects of plausible cyber attacks as highlighted in a study by Oxford University, namely: assets targeted, threat actors, outcomes of the attack and attack vectors. Security policies and mechanisms should aim to protect the assemblage of assets - cameras, networks, databases and analytics tools - that constitute the smart CCTV infrastructure.

For example, neighbourhood watch groups could be alert for signs of suspicious activities (for example, drive-by hacking) in the proximity of police cameras and network infrastructure, in addition to classic neighbourhood crimes.

Security agencies' policies on the implementation of smart CCTVs should factor in other critical factors, including interoperability with mission-critical systems such as criminal intelligence databases, and command, control and communications systems; information-sharing between agencies; and addressing the unintended and unexpected implications, such as public expectations of law enforcement standards with respect to police presence and response.

In a 2013 Federal Bureau of Investigation bulletin article on predictive policing, the Santa Cruz Police Department emphasised that technology could supplement but never supplant the innate attributes of effective law enforcement, such as good investigative instincts, human intelligence and community engagement. Undaunted, adversaries might adapt their tradecraft to outsmart electronic surveillance.

For example, the Bastille Day attack in Nice occurred despite the city being known as the "CCTV capital" of France.

At present, while the AI in smart CCTVs can highlight potential security concerns, it cannot yet perform investigative tasks, such as assessing the intent and capability of suspects. Over-reliance on technology might also affect officers' alertness to danger and regularity of face-to-face interactions with people on the streets. Thus, an intermediate knowledge of smart technology is now critical in the skillset of officers to make them both tech- and street-savvy.

Smart CCTVs will have a critical role in the coming years in securing cities and in homeland security. Their proliferation would raise privacy concerns and their omnipresence could inadvertently create the illusion of gated communities and increase complacency in personal security.

Subject to a city's socio-cultural context, legislation such as the Data Protection Act in the United Kingdom would help to assuage privacy concerns by regulating the responsible use of CCTVs. A healthy community partnership, such as in Singapore, would help the public to acknowledge the necessity of smart CCTVs for the collective good, and that both community vigilance and smart CCTVs are concomitant and essential aspects of enhanced crime-prevention and security strategies.

  • The writer is a research fellow with the Homeland Defence Programme at the Centre of Excellence for National Security (CENS), a unit of the S. Rajaratnam School of International Studies (RSIS), Nanyang Technological University.
  • This article first appeared in RSIS Commentary.
A version of this article appeared in the print edition of The Straits Times on December 24, 2016, with the headline 'Smart CCTVs: Third eye of secure cities'. Subscribe