Event Data’s Role in IoT: How to bring event data to life for better IoT analytics and customer insights


Organizations are making the leap from traditional transactional data sets to the expansive world of event data sets, which can come from IoT-enabled devices. Consumers generate event data from step trackers, Internet-enabled thermostats, and autonomous driving cars. Businesses track events from the manufacturing floors, the fulfillment center or warehouse, and from their IoT-enabled fleet vehicles, such as container ships, trucks, and (soon) delivery drones. There are also events generated by mobile apps and online portals. From these events, organizations collect information on where and how customers, partners, and suppliers interact with informational, ordering, payment, and delivery systems. There are also events generated from real-time applications and security platforms.

Enabling Architectures

IT departments and technologists have myriad options to create, manage, and monitor event data flows. One of the most common is Apache Kafka, which directs message streams from “publishers,” such as enterprise applications or IoT devices, to “consumers,” such as data lake/big data targets or other streaming pipelines for analysis and correlation. While Apache Kafka is popular, there are other streaming platforms, including Amazons Kinesis, Azures Event Hub, and MapR Event Data Streams. IT stakeholders also might use change data capture technologies in their solutions, such as Attunity Replicate, to stream live source database transaction information to a similar set of “consumers” or targets.

Attunity Replicate/IoT Diagram

Focus on Customer, Partner, and Supplier Data for Better Value

The initial stage of information and analysis is about the events. How many steps and levels of quality of production are there? What’s the location of containers or packages? What are the numbers of orders from mobile and online? How about the timing of payments made? However, the next level of value in event data is to break free from the date, time, and specifics of the events themselves and link those events with customers, partners, and suppliers who initiate those events and understand their behaviors. Are visits spread across the entire customer base, or simply focused on a relatively small number? How do the inputs provided by particular suppliers impact production quality? Which partners are providing the most value to your supply chain?

Add Contextual Info for Better Customer Behavior and Trending Insights

To make the jump from a “simple” event to one that has the proper context, organizations need to pull together information from across their data landscape. Most of this contextual information on customers, partners, suppliers, inventories, and sales resides in enterprise applications such as SAP and other enterprise resource planning, supply chain management, and customer relationship management platforms. This contextual information takes events consisting of date, time, product ID, customer ID, and other information (e.g., geolocation and sensor reading data) and turns them into steps along a process “journey”. For example, adding contextual information to event data details how a customer moves from initial contact to the products they evaluated and ultimately selected (or, more importantly, didn’t purchase), and how often they come back to make repeat or associated purchases.

Leveraging Modern Data Integration

With the scale and breadth of not only event data, but also the information within enterprise applications, organizations cannot use traditional data management and integration practices, they can take the strategic best “patterns” of those approaches. But, they really need to adopt the tactics and methodologies of modern data integration. This means being able to coordinate the movement of data across platforms to understand where data and information reside. It means using automated techniques to hasten the deployment of new integrations and optimize the quality and speed of implementation of existing integrations. Also, organizations must have a monitoring and management facility to understand how all these moving parts are working. They also need to know whether data management and engineering teams need to take action to ensure event data and contextual information should be combined to best meet the needs of the organization.

To learn more about enabling 360-degree business insights, check out the recent Enabling 360-degree Business Insights with SAP Data on-demand webinar presented by experts from EMA and Attunity.

About the Author

This article was written by John Myers, Managing Research Director at EMA.

Dev Tool:

Request: blog/event-datas-role-in-iot
Matched Rewrite Rule: blog/([^/]+)(?:/([0-9]+))?/?$
Matched Rewrite Query: name=event-datas-role-in-iot&page=
Loaded Template: single.php