5 of the best data analytics platforms that use the edge

The ability to use edge computing to optimize data processing and develop real time analytics ability is available to all organisations with the help of these platforms.

IBM Watson
IBM

There will be an estimated 20.4 billion connected things by 2020, placing an extreme pressure on network systems. The vast amount of data generated by these smart sensors and devices is currently largely being processed in the cloud or data centres, which will not be adequate for long. The necessity is arising for another way to handle the quantities of data, and that is edge computing.

The edge computing model comprises sensors and connected devices to communicate with a gateway device, instead of the cloud or server, which processes the information. This technology is better suited for gathering and processing IoT generated data as it can deliver almost instantaneous analysis. That is important for industrial applications that involve making process changes based on recorded values. It is also has economic implications, as it is less expensive regarding data management, and by utilizing a gateway device, it negates the cloud network and results in less constricted networks. Another advantage, is because of its standalone architecture, the network will remain operational even if there is a problem with one of the device.

There are various platforms that can be used to optimize edge-based data analytics. Here are five CIO’s need to be aware of.

1. Azure stream

Microsoft’s IoT platform delivers a solution offering a real-time analytical solution that is popular with the manufacturing industry. It can handle complex event processing (CEP) pipelines readily, and seamlessly integrates the IoT ecosystem. It leverages an SQL analogue with temporal logic support embedded to provide capabilities for machine learning and predictive analyses.

The program can be combined with other Azure analytics to produce to provide a powerful interface for data processing. This product claims to save developer resources as it has superior debugging and job monitoring abilities. Fully customizable, real-time dashboards help users utilize generated data to provide insights and predictions to use for effective decision making. The Azure platform is appropriate for critical workloads and easily scalable, providing a useful edge computing tool for viable data analytics. 

2. IBM Watson

IBM’s IoT analytics framework is an extension of the IoT Watson platform, which delivers a cognitive analytical model to help users understand scenarios. Built on Apache Edgent, IBM provides processing, analytical power and machine learning at the edge of the network. This cloud and edge analytics model allows optimal data control between edge devices and the cloud, allowing differential data processing. Specifically regarding which data is processed at the edge and which is sent to the cloud.

This type of data filtering at the edge is beneficial as the gateway device will know how to process and respond to instructions from the platform. The IBM model effectively provides a valuable framework for an industrial setting, as responses are pre-defined and instantaneous.

3. Cisco Connected Steaming Analytics

The drawcard for this platform is its streaming function whereby it streams data from various sources to collaborate and glean useful insights in a big data model. There are various architecture options which are both horizontally scalable and allow local data processing. The result is a high performance platform that provides instantaneous governance and actionable insights.

Cisco connected streaming analytics provides commercial value in the form of real time decisions and knowledge that can be exploited to optimize business processes. The advantage of this platform is its high velocity data streaming from multiple sources which can be used by various groups within an organization to facilitate a multi-disciplinary framework within the corporation. The real-time multidimensional analysis also ensures efficient use of resources.

It is built specifically for edge computing and is suited to continuous monitoring of live data streams and claims to provide the best in class variation of adaptors. This makes it easy to integrate into current systems, as well as interface with variable data sources.

4. Oracle Edge Analytics

The Oracle edge product filters, correlates and processes data on embedded devices to support downstream applications, providing real-time analytics. This platform is best suited to event processing applications that require intelligent compliance and responses.

The key features of the Oracle platform are its high processing speed and real time data capture capabilities which support high throughput and eliminates latent processing.

5. Intel analytics toolkit

Intel’s IoT platform comprises data collection resources, as well as analytical capabilities from smart sensors. The advantage of this system is that it can be used as an entry level analytics ecosystem without an expensive and large-scale outlay, as it requires little storage and processing capacity. It also harnesses the abilities of Apache Hadoop to capture data and leverages simpler analytical software design to provide a wide range of capabilities and to bring big data solutions to organizations of all sizes and budgets.

Intel has combined entity-based machine learning with end-to-end graphical processing that involve algorithms for elucidating correlations in big data. Its modular architecture allows easy extension and integration with further analytical applications.

Edge computing provides a useful platform to optimize data processing within devices to provide data driven organizations with the ability to optimize their analytic capabilities and draw useful inferences in a real-time, cost-effective and powerful way.

This article is published as part of the IDG Contributor Network. Want to Join?

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Now read: Getting grounded in IoT