How Edge Computing Affects These 3 Application Technologies

by Brad Johnson, on Apr 17, 2018 11:19:16 AM

As organizations adopt Internet of Things (IoT) technologies in pursuit of more productive and cost-effective operations, it’s inevitable that we’ll continue seeing a shift from the cloud as a center of computing to more distributed approaches which leverage available compute at the edge. As compute becomes cheaper, thanks to Moore’s Law, edge devices are becoming increasingly capable of performing complex analytics locally. Furthermore, applications are becoming more distributed, and must support increasingly diverse sets of devices at the network edge. Edge computing technologies will enable, and rely on, innovations elsewhere in the application stack.

These three application technologies will help to usher in the future of edge computing, and in return edge computing will help unlock new use cases:

  • Streaming/Real-time Analytics will form a foundational building block of successful edge computing applications. As streaming edge data is generated, streaming analytics can power real-time data filtering, local control, and deliver insights via APIs to local user interfaces (UIs) and downstream applications. Edge computing will enable streaming analytics technologies to be deployed closer to the data source, significantly reducing incurred network latency and achieving further efficiency benefits.
  • Digital Twins are virtual representations of real-world objects, consisting of data from and other contextual information about that object. Digital twins maintain consistent state with the objects they represent, ensuring consistency throughout an application. Digital twins simplify application development by providing a uniform way of addressing otherwise heterogenous objects in a system, abstracting away the notion of an individual device from the rest of the application. The  distributed actor model (used in Lightbend’s Akka, Microsoft’s Orleans, etc.) provides an application pattern for building digital twins, providing a means of ensuring eventual consistency throughout an edge application.
  • Machine Learning/Artificial Intelligence technologies can be deployed at the edge to perform a variety of tasks. Self-training machine learning algorithms can intelligently filter “gray” data at the edge, greatly reducing the volume of data which must be transmitted over local networks or stored later. Deep Neural Networks deployed at the edge can observe data as it’s created, continuously testing hypotheses to form highly accurate real-time predictions. This will enable Machine Learning to power local, real-time applications, continuously training on streaming sensor data at the edge, to inform maintenance, production, ERP and other applications.

Learn More

Learn how SWIM EDX uses edge intelligence to deliver real-time insights from the dark data generated by connected enterprise systems.

Topics:Machine LearningStateful ApplicationsSWIM SoftwareEdge AnalyticsDigital TwinSWIM AIEdge ComputingFog ComputingSWIM EDX