Nebula AI, a provider of intelligent monitoring solutions for industrial IoT deployments, today announced the launch of Nebula Sense. This is an artificial intelligence (AI) technology that monitors industrial IoT deployments to detect and automatically mitigate anomalies in real-time. Using three layers of machine learning algorithms that are trained to recognize different types of devices and anomalies, Nebula AI can accurately identify abnormalities without requiring any human intervention or additional training.
Three layers of machine learning
- The first layer learns to recognize the different devices and anomalies that can exist in industrial IoT deployment.
- The second layer learns how to detect and automatically mitigate the anomalies.
- The third layer, which is the most important, learns how not to need any human input or additional training for these anomalies.
The IoT is a rapidly expanding area of technology that connects machines to the internet, and this connectivity leads to new vulnerabilities. A report from ICS-CERT found that in 2017 there were over 800 reported cyber incidents involving industrial control systems devices (ICS). In 2018, there have been reports of successful or attempted intrusions into various energy company networks. A recent attack on an energy company in Ukraine cut power to nearly half the country for hours, and this was not a one-off event.
Nebula Sense is designed so that any data collected can be automatically analyzed by AI algorithms running in real time without human intervention or training required. The application of three layers of machine learning algorithms that are trained to recognize different types of devices and anomalies means the technology can accurately identify abnormalities without any human intervention.
Nebula Sense is available for free on Github at: nebula-iot/nebula-sense/.
Narrow-band IoT, or NB-IoT, is a new air interface aimed at low data-rate transmission for devices with time constraints. It is designed to operate in licensed spectrum, and it works well in areas with less than reliable coverage due to its low power requirement.