12 Edge analytics

 

This chapter covers

  • Using Pulsar for edge computing
  • Using Pulsar to perform edge analytics
  • Performing anomaly detection on the edge using Pulsar Functions
  • Performing statistical analytics on the edge using Pulsar Functions

If you are like most people, when you hear the term the Internet of Things (IoT), you tend to think of smart thermostats, internet-connected refrigerators, or personal data assistants, such as Alexa. While these consumer-oriented IoT devices tend to get a lot of attention, there is a subset of IoT called the industrial internet of things (IIoT), which focuses on the use of sensors that are connected to machinery and vehicles within the transport, energy, and industrial sectors. Companies use the information collected from sensors that are physically embedded inside industrial equipment to monitor, automate, and predict all kinds of industrial processes and outcomes.

12.1 IIoT architecture

12.1.1 The perception and reaction layer

12.1.2 The transportation layer

12.1.3 The data processing layer

12.2 A Pulsar-based processing layer

12.3 Edge analytics

12.3.1 Telemetric data

12.3.2 Univariate and multivariate

12.4 Univariate analysis

12.4.1 Noise reduction

12.4.2 Statistical analysis

12.4.3 Approximation

12.5 Multivariate analysis

12.5.1 Creating a bidirectional messaging mesh

12.5.2 Multivariate dataset construction

12.6 Beyond the book

Summary