To achieve maximum benefit, city planners, governments, and private businesses need a platform that connects all the dots. Technologies that can work together to improve insights for a city like video surveillance, access control, automatic license plate recognition, cloud services, operational decision support, and investigative case management are typically in place already.
“Take city traffic
, for example,” said Giovanni Gaccione, Justice & Public Safety Practice Leader at Genetec
. “Video surveillance cameras, automatic license plate recognition, and analytics can work together to improve traffic. Law enforcement can identify incidents quicker – which means they can respond and clear the roadways faster.”
When situations arise, police and local organizations can use technology to communicate detours which results in better traffic flow and happier citizens. These solutions can even work with transit providers like subways or buses and can be configured to notify riders in real-time which route to choose based on traffic conditions ahead. For all this to translate to predictive security, a few more elements should be in place.
Components required for predictive analytics
Safe cities are locations where tackling urban crime involves more than just using a police force. Peter Matuchniak, CTO of Maxxess Systems, in developing any city solution there are three vital components. These are external and public domain alerts, which are generally available information such as news bulletins, weather reports, system alerts like data from facial recognition, access control systems, and human intelligence which is citizens or team members reporting issues directly using their phones.
Brian Schwab, Founder and Principal Consultant at S3SDC
elaborated on this further, giving details of key high-level elements of safe city
System of interoperable sensors
These should be integrated onto a shared network and include
- Physical sensors such as acoustic gunshot detectors, video cameras, and CBRNE and fire sensors to detect and map fire or chemical agent releases in a monitored area.
- Cyber sensors such as social media monitors, web-based sensors using risk models, and IoT sensor monitoring and communications gateway (Machine-to-Machine (M2M) network) that allows sensors to communicate and ultimately connect back to a central system for live feeds of what’s happening on the ground.
Situational awareness components
These include solutions where real-time information from weather, traffic, resource locations coming from a variety of integrated sensors, cameras, and other security equipment is mapped to give a hyper-accurate picture of the current operating environment. Components include ArcGIS platform for fusing the data within a geographic overlay and event processing platform.
Video data & analytics
This is where data from multiple systems is integrated and the output is processed with video analytics, LPR
, facial recognition
, behavior analysis, and other processing software to identify potential threats. The cameras serve as the eyes of the system in public spaces, provide data for forensic analysis, suspect identification and movement tracking and allow for detection and assessment of non-normal activities.
System components include, but are not limited to smart cameras at the edge, flexible cloud infrastructure/centralized server, video management systems, and video-based metadata
Mobile broadband and cloud technologies
Mobile broadband and Cloud technology are necessary to support the IoT-based information and communication technology (ICT) systems forming the underlying base of safe city infrastructure. This permits rapid communication of multiple data streams to a central location, where it can be processed, analyzed, distributed and actioned by security and safety personnel across the city.
Automated processes refer to solutions where information is streaming in from multiple points and is gathered then collated. Alerts are registered, and the correct response is instigated with the use of alarms, alerts and other generated protocols.
To make the most of predictive analytic solutions, different government agencies should share intelligence, operational procedures and planning.
A work in progress
Andrea Sorri, Director, Business Development Government, City Surveillance and Critical Infrastructures at Axis Communications
warns that predictive analytics is still in its infancy.
“The ability to predict what is likely to happen in the future is based on machine learning, and effective machine learning requires a lot of data,” Sorri said. “So, this is the first requirement: a volume of data sufficient enough to allow computers to learn and, from that, predict.”
The volume and variety of data needed to allow for predictive analytics also demand that camera data alone – though a vital component - is not enough, Sorri added. It needs to be combined with data from other sensors, including thermal, radar, pollution and air quality monitoring, smoke detectors, sensors to measure rising water levels and many others. Combined, these inputs can start to accurately predict future events and remedial action taken.