Automating Surveillance Workflows

An incorrectly set up surveillance system can be annoying. Pinging you hundreds of times a day with irrelevant information quickly becomes more of a hindrance than a help. “Alert fatigue” is a real issue.

Older surveillance cameras rely mainly on zones (a user-determined detection area) and simple motion detection to determine if an event is worthy of your attention. Configuring this can be difficult and frustrating at the best of times. If you adjust your motion detection to be too sensitive, you are quickly inundated with alerts, turning the sensitivity down too low, and you’ll miss important events. Adding to the frustration is the fact that every individual camera model will have to be adjusted with a unique value.

These configuration issues have recently been solved using more complex image processing, the best solutions now incorporating artificial intelligence (AI). New AI infused visual models can be trained to detect people and vehicles, even individual faces and license plate numbers. The advantage of this is the ability to create a set of rules that notifications  (SMS messages and emails) must then follow. For example, you could set notifications only to be sent when a person enters a specified area at a certain time (not a dog, a squirrel a cloud moving across the sky or a jiggling spiders on its web). A correctly configured system would then only send you notifications of the things you care about. You, no doubt, already have enough demands for your attention.

A correctly configured, modern surveillance system is imperative to keeping your signal to noise ratio at sane levels.

Remote Equipment Monitoring