LATEST THINKING

Find your IOT edge

Bringing analytics to data at the edge of IOT

BRINGING ANALYTICS TO DATA AT THE EDGE OF THE IOT

As Internet of Things (IoT) deployments become increasing complex and business increasingly data-driven, a challenge arises: How to combine scalability with manageability?

The challenge is exacerbated by escalating volumes of data and by network limitations in bandwidth, latency and connectivity. Bringing the data to the analytics is no longer sufficient. It’s time to bring the analytics to the data.

Accenture Labs has created an edge analytics framework that addresses these challenges without re-inventing the wheel. A key goal: Avoid wholesale changes to an organization’s IT environment.

We aim to work with existing environments with a varied applications, models and hardware infrastructure, and, simultaneously, pave a way forward to take advantage of a technology refresh.

How does it work? Read on for details.

arrow

THE CLOUD IS MOVING TO THE “FOG”

We distinguish edge analytics from edge computing in the need to run a variety of machine learning, predictive and prescriptive models that require a more powerful computing environment with specific CPU, memory and storage requirements.

Today, not every edge device is suitable for providing such an environment to train, run and re-train these models. In the “fog” layer, edge gateways typically serve as the primary connectivity points in the field for various IoT devices and industrial equipment, including legacy and passive sensors.

The small footprint means they can be deployed more closely to the IoT devices in the physical environment, and offer basic levels of computing, storage and network capabilities suitable for analytics.

TRIFECTA OF CAPABILITIES

An edge analytics platform must manage the deployment of the models associated with the edge device in the fog, and, in turn, push it out to the device. Specifically, an edge analytics platform enables you to:

Device Expert

Centrally develop, train and manage analytics models in the cloud by leveraging data from a global population of edge devices

Application Expert

Deploy these models to execute in the fog to take advantage of unfiltered, high-fidelity data and low latency response times, and

Field Engineer

Seamlessly coordinate with the cloud-based platform for the models to adapt to the specific dynamics of the local environment.

Combined, these three features enable the enterprise to provide insights where they are relevant and to drive immediate outcomes.

ENABLING THE “DIGITAL TWIN”—BRIDGING THE EDGE AND THE CLOUD

As you continuously deploy and train these models on the edge, you can customize assets at a finer level, while the cloud helps combine insights from data populations across devices.

This paired approach, bridging the governance of the cloud to the edge, leads to a virtuous cycle for creating the self-optimizing models essential to digital “twins.” Our comprehensive approach helps enable the complex models digital twins require.

A twin simultaneously resides on multiple instances of cloud nodes and edge devices and requires coordination to continue learning and improving over time—creating a feedback loop that results in a continuously evolving living service.

Read the full report for more on the opportunities in edge analytics.

"Now, we can handle edge analytics at scale and tap into high-fidelity data and contextual processing directly at the edge. Rather than bringing the data to analytics in the cloud, it’s time to bring analytics to the data at the edge."

SUGGESTED CONTENT