In a busy mall, someone drops a bag and walks away. How long will it be before security staff notice? And when they do, how long will it take police to identify the person who dropped the bag if it turns out to contain something dangerous, such as a bomb?
With a video surveillance system, security personnel likely would notice the bag faster than if they depended entirely on guards making their rounds. However, it could take police hours to scour the footage to pick out the person who left it. The longer law enforcement officers spend on a forensic investigation, the more time the suspect has to get away.
Next-generation video analytics software using artificial intelligence could automatically alert guards to the dropped bag and to other incidents, such as fights. AI enables faster responses and lower overhead costs rather than relying on a small army of humans to monitor every camera feed.
Some software can even identify specific people in a crowd. In this scenario, the software would notice the dropped bag, quickly go back to the recording to find out who left it and then present a screenshot to security staff or police. That speed could allow security to cordon off the property before the suspect escapes.
Gain Hindsight, Insight, Foresight
Video analytics software is a key enabler for hindsight, insight and foresight in surveillance. These principles can deliver astounding benefits for organizations such as retailers, schools, governments and others that want to maximize the value of their video surveillance systems.
- Hindsight accelerates forensics. For example, a retailer may use analytics software to determine how many shoppers — sorted by factors such as gender, age and sentiment — stopped to watch an ad on a digital sign or peruse a display.
- Insight provides immediate situational awareness. This awareness can help organizations improve security and efficiency. For example, video analytics can identify known and unknown persons entering a school to make sure unknown individuals check in with the office.
- Foresight applies the lessons learned from hindsight. Analytics software could alert a city, for example, to emerging traffic patterns that indicate where lanes or speed bumps should be added to reduce congestion and improve safety.
Scale Up Efficiently and Effectively
More cameras mean more opportunities for hindsight, insight and foresight — and all of the safety, productivity, efficiency and other benefits they enable. One new use case is drone-based cameras, which provide enormous flexibility. For example, law enforcement agencies can fly them over crime hot spots to keep an eye out for potential problems or over major sporting events for crowd control.
When scaling up a surveillance system, two common challenges involve personnel and connectivity. Often, organizations don’t have the necessary workers to keep an eye on and analyze surveillance footage. Further, establishing network links between cameras and a centralized platform can be challenging. Analytics can help organizations address both obstacles.
AI-powered analytics minimize the number of staff required to monitor feeds; for instance, by alerting humans only when a camera detects certain events. That saves money. Such analytics solutions also reduce bandwidth costs by doing their work at the network edge, even in the camera itself. So instead of continuously sending every feed to a central monitoring location, the software turns on each camera’s backhaul only when the AI detects something that a human needs to see right away (insight) or that should be recorded (hindsight and foresight).
Drones and video analytics are just two examples of how far and how quickly video surveillance technology advanced over the past few years. Organizations that navigate these new options effectively can make video surveillance a key part of their digital transformation.