Autonomous transportation is still futuristic, but freight trucking is pulling ahead Autonomous vehicles are quickly becoming the next step for the majority of car and technology companies alike. There is still a long way to go before this mode of transportation becomes the norm, but it may arrive [...]
Smart city technologies can reduce crime by 30-40% by leveraging real-time monitoring systems to automatically send alerts to law enforcement.
Smart cities are leveraging emerging technologies to improve the way of life in urban regions. These, technologies, which offer an enticing range of benefits including better traffic management, better healthcare facilities, and heightened connectivity, can also help with reducing crime, thereby creating safer cities.
According to a McKinsey Global Institute (MGI) report, smart city technology can reduce crime by 30-40% thanks to predictive policing, real-time crime mapping, and home security systems. Furthermore, it can accelerate emergency response times by 20-35% and reduce fatalities from homicides, road traffic, and fires by 8-10%.
One example of a smart city crime-fighting project is the “Gunfire locator” installed across the city of Oakland in California. These locators are fitted with microphones that can detect a gunshot and then triangulate the position to send the nearest patrol car. The system, which has been in place since 2012, has been able to reduce firearm incidents by 66%.
Meanwhile, in Southeast Asia, due to improved spending power, digital literacy, and smartphone penetration, many cities are adopting smart solutions. According to a report, smart cities can save around 5,000 lives that are lost every year to traffic accidents, fires, and homicides by optimizing for mobility, crime prevention, and better emergency response.
Analytic CCTVs for real-time monitoring
“The objective of smart cities is to be able to harness data, including from CCTVs, so that it can be used in various ways, including for security, safety, [and] operational efficiency,” Abhijit Shanbhag, founder and CEO of Graymatics, tells Jumpstart.
A leading cognitive media processing company, Graymatics provides AI-powered solutions for multiple sectors including security and surveillance, digital marketing, telecommunications, and the Internet of Things.
An important segment within the smart city sector is analytic CCTV, in which CCTV systems perform real-time monitoring to automatically send alerts to law enforcement agencies whenever any kind of criminal activity or anything related to safety is detected.
“Graymatics has developed a very [richly] featured video AI platform which can ride through all the idiosyncrasies of CCTV and be able to analyze all the CCTV video feeds to an astonishing level of depth,” says Shanbhag.
For instance, in the context of the pandemic, CCTV analytics can detect whether people in the vicinity are violating social distancing rules or not wearing masks. In scenarios related to crime, smart CCTV can detect if a group of people is fighting among themselves or if a person steals a wallet from someone walking by. In terms of safety, it can tell if a person is riding a motorcycle without a helmet.
“All of those actions will be captured [and] obtained by Graymatics platform by analyzing the CCTV feeds,” Shanbhag says. Whenever the startup’s system detects a crime, he adds, law enforcement will be automatically alerted.
In order to effectively act on these alerts, law enforcement officers will have to be properly deployed in different parts of the city. However, when the smart city is first deployed, law enforcement may not know how best to distribute officers. This is where data analytics is able to play a role in resource optimization.
“We provide a very detailed dashboard, indicating various insights as to what time [and] in what region, does a specific kind of infraction happen,” Shanbhag says.
These insights can be utilized to predict where some of these infractions are likely to happen during the day, or where a specific type of infraction is more likely to occur. Once these details are known, the law enforcement officers can be deployed in a far more structured and action-oriented manner.
In addition to CCTV, there are sensors that can be deployed to detect certain kinds of activities. For example, weight sensors can detect if a large number of people gather in crime-prone areas – useful information given that according to Shanbhag, crowds raise the likelihood of a crime occurring.
Ultimately, Shanbhag says, as vision is the main sensor to detect crime, “the CCTVs will be the best approach to detect crimes within smart cities.” However, he does note that this doesn’t protect against cybercrime.
Further, video evidence plays a crucial role in the admission of AI-identified evidence of wrongdoing in court.
“Society is only beginning to embrace AI and there will always be some level of skepticism that this is a whole new technology which is coming in, and how accurate could it be,” says Shanbhag.
“So, it’s very important that when AI detects these alerts, an appropriate video clip is also correspondingly stored relating to that alert,” he explains. The existence of the video clips can then prove that there really was an infraction, as opposed to a false alert sent by the AI system.
Tackling the limitations of analytic CCTVs
The main limitations of these CCTVs are two-fold: one, the heights at which they are installed, and two, their use at night.
“What has commonly been considered a limitation with CCTV is that they may be placed five and a half to six meters high. At that height, a number of companies have indicated that it’s difficult or even impossible to be able to carry out facial recognition and different kinds of people-related analytics,” Shanbhag says.
“The CCTVs have to be also placed appropriately in strategic locations so that it can be properly leveraged and exploited. Because finally, any crime will be detected only if it’s within the eye view of the CCTV,” he adds.
The second limitation is that analytics cannot be effectively carried out in limited lighting, which makes nighttime crime-spotting a problem. One solution, Shanbhag says, is integrating lighting, such as LED lighting, into the CCTV. Alternatively, infrared CCTVs can be used, but these are much more expensive.
Besides these two major issues, another limitation is how well the CCTV systems perform under weather conditions like haze or rain. These, Shanbhag says, can also be controlled for using powerful algorithms.
When it comes to smart cities, a major area of concern is the privacy of the residents. What would happen to privacy in a city constantly monitored by cameras and sensors?
“The privacy issue comes in when the data obtained from all these sensors is being utilized in ways that could compromise people’s privacy,” says Shanbhag.
For instance, when facial recognition is enabled, it will detect that a particular person went to the mall, accompanied by two others – a clear violation of privacy.
“[However], if you disable some of the elements – let’s say facial recognition should be limited to detecting only the blacklisted people – [then] it’s all about promoting community safety. Then, privacy will not be an issue,” Shanbhag explains.
That said, the command control centers and servers where the video feeds are stored could be vulnerable to cyber threats.
In 2018, the U.S. city of Atlanta faced a massive ransomware attack that rendered its City Hall powerless for five days, disrupted law enforcement, payment of parking tickets, processing of court cases, and more, highlighting the need for smart city security.
Emphasizing the importance of security, Shanbhag says, “Within the command control center where all the video feeds and all the other sensor feeds are placed, there has to be a very high level of security, so as to be able to block any kind of illegal access.”
These security measures, he adds, are largely centered around different levels of encryption, authentication, and network security.
Smart cities in India
The requirement for a smart city or a safe city is far more in developing countries like India than in developed countries, where traffic flows in a more ordered manner or has a more mature legal structure, says Shanbhag.
In India, a very populous country, “things tend to be less structured in terms of traffic, some of the legal rules, and how the society’s mindset is,” he says. This is where the many benefits of smart cities can be leveraged in unique ways.
“There is a lot more room to improve on the orderliness of the traffic, [and] in detecting various crimes which would otherwise very likely might go unreported,” Shanbhag adds.
However, when it comes to deploying smart cities in India, one of the major challenges is finance. From setting up a data center with appropriate amounts of computing hardware to deploying the CCTV cameras, everything requires a budget.
“The benefits are clearly a lot higher for a developing country, especially a populous country. But at the same time, the required budget will be the limitation,” he notes.
However, in India, both the central and state governments have earmarked various budgets for smart cities. In 2015, the Government of India launched the Smart Cities Mission to “drive economic growth and improve the quality of life of people” through local development and the adoption of various technologies.
For this, the central government has offered financial aid of INR 480 billion (US$6.5 billion) over five years at an average INR 1 billion (US$13.7 million) per city per year. Along with this, an equal amount will also have to be contributed by the respective state or union territories.
The mission will cover 100 cities, each of which has to focus on core infrastructure elements including sustainable environment, adequate water supply, assured electricity supply, robust IT connectivity and digitalization, and the safety and security of citizens, among others.
However, the project, which was slated to be completed by 2020, is way behind. From 2015-2019, only half of the designated amount was put toward the project.
Meanwhile, there are several private entities from across the globe that are looking into investing in smart cities in India, Shanbhag says. This, he adds, may be a positive force in rolling out smart city technologies across the nation.
The road ahead
While there have been some smart cities-related developments happening across major cities in Southeast Asia, Shanbhag says this development is only the tip of the iceberg.
“There will be lot more happening over the next several months, and next several years for cities to really evolve to be considered [smart],” he adds.
Shanbhag says that there will be an ecosystem of companies providing pilots to the cities, perhaps free of cost, for a certain period of time.
“The city can identify all the benefits that it accrues from this technology in maybe over six months. Based on that, it may provide a contract on a nomination basis,” he adds.
Additionally, Shanbhag predicts that the industry may see outcome-based pricing in the next few months. Under this, the cities which have deployed smart city technology as a pilot can identify how much they can earn from the collection of traffic violation tickets and other minor infractions. This may make possible a certain degree of revenue sharing with the technology vendors.
Meanwhile, in terms of emerging technologies, leveraging audio in addition to CCTV video analytics is one area that will see momentum in the coming years. For example, while CCTVs have a definite scope of view, many are also moveable.
“Let’s say there is someone who is screaming for help outside of the scope of view of the CCTV. The ideal scenario is utilizing the audio analytics to steer the camera in those directions where the crime is likely happening,” Shanbhag says.
Robots equipped with cameras that can track crime based on various sensor-related data is another technology that Shanbhag says we can expect.
Smart cities are primed to re-envision urban living. For smart city initiatives to kickstart, the government, startups, law enforcement agencies, and citizens have to work in tandem, all the while ensuring the privacy and security of the data collected.
Header image courtesy of Graymatics