Customer Stories

Going with the flow: exploring real-world use cases for real-time streaming analytics

Share this post:

If you were to make a list of the most-hyped topics in enterprise technology right now, it’s likely that the Internet of Things (IoT), big data and event-driven applications would be near the top of it. In many cases, these types of solutions share a common factor: the need to capture, analyze and act on fast-moving, high-volume streams of data—whether that data is generated by IoT sensors, user activity on the web, or traditional transactional data such as trading on financial markets.

At a high level, it’s clear that many organizations are getting excited about the broad possibilities of these types of stream computing solutions. However, when you get down to the specifics, they often find it difficult to identify or articulate any practical real-world use cases.

In many cases, there’s a “wait and see” attitude—companies are sure that valuable use cases exist, but they won’t take the first step until a really solid business case emerges. This is a sensible, low-risk approach—but it also means they are forgoing the chance to innovate, and potentially handing competitive advantage to their more agile and disruptive rivals.

Perhaps what’s needed is a little inspiration: let’s take a look at some of the early adopters that have already successfully built these kinds of applications, and explore the potential for future developments in various industries and functional areas.

First, let’s consider the IT and telecom industries. As networks and infrastructures become ever larger and more complex, managing them efficiently and maintaining system availability becomes too great a challenge for human system administrators to manage unaided. Stream processing technology can help by capturing, analyzing and ingesting millions of events from device logs across the network, highlighting problems, and facilitating rapid root cause analysis.

For example, Cerner, a company that provides IT services for the healthcare industry, is using a combination of IBM Streams and Apache Kafka to monitor performance abnormalities across 1.2 billion real-time monitoring system (RTMS) daily timers, and apply an advanced algorithm to gain faster, more detailed insight into system performance. As a result, its IT team can react much more quickly when a system shows signs of performance degradation, helping to keep its clients’ vital healthcare systems online 24/7.

Similarly, GFM Integration, a systems integrator that works with the communications industry, built a solution that captures time-series data from telecom network devices and analyzes it in seconds to analyze patterns in call data records (CDRs) to pinpoint faults or areas of poor network coverage. This not only helps to find and fix problems faster; it also supports better network planning by showing companies how network usage is evolving over time.

Network monitoring is not the only stream processing use case in the telecom sector: there is also considerable potential for the technology to improve the customer experience. For instance, Verizon has been working on a call center solution that converts customers’ speech to text, analyzes the text to identify the key topics, and searches for relevant information in the company’s knowledge base within seconds. The aim is to give call center agents the information they need to solve the customer’s problem within a single phone call, potentially increasing satisfaction levels while reducing support costs.

Of course, this type of solution need not be considered specific to the telecom industry: similar solutions could be built with IBM Streams to support customer service teams in any sector that needs to provide support for large numbers of customers, from banking and insurance to utilities and government.

So far, we’ve looked mostly at solutions that augment existing systems with near-real time analytics—but what about the opportunities presented by building new networks of connected devices? For early success stories in this area, look no further than the healthcare industry: IoT-enabled medical devices present some of the most exciting use cases, and the clearest indications of value in terms of increasing efficiency and improving patient care.

For example, the Irish Centre for Fetal and Neonatal Translational Research (INFANT) is using IBM Streams to capture brainwave data from electroencephalograms and pass it through machine learning algorithms to detect seizures in newborn babies—a task that traditionally depends on expert interpretation by highly skilled neurophysiologists. When a seizure is detected, the system alerts physicians immediately, helping them intervene to relieve the symptoms and reduce the risk of long-term damage.

Similarly, the Children’s Hospital of Philadelphia (CHOP) is working on a project that will use IBM Streams to monitor data from ventilation monitors and use machine learning to help physicians check whether a patient has been intubated successfully—often a significant challenge in pediatric care, since children’s airways are much smaller than adults’. By reducing the need for more invasive methods of checking a patient’s intubation status (such as X-rays), the solution aims to improve patient safety and avoid delays in surgery.

Finally, Clemetric, a solution provider for the healthcare industry, has built a general-purpose streaming analytics platform called Q2Care, which harnesses IBM Streams to gather physiological data from electrocardiograms, pulse oximeters, pulmonary artery catheters and a wide range of medical devices, and present it in real-time dashboards that help physicians understand a patient’s condition at a glance. The ability to capture and correlate readings over time also helps to identify and detect patterns that could potentially predict when a patient’s condition is likely to worsen in the near future.

Hopefully, these examples should have given you some ideas about how stream processing could help your organization seize the opportunities presented by the IoT, or gain new value from your existing data by analyzing it in flight, rather than just retrospectively.

For more information on use cases for streaming analytics, take a look at the IBM white paper, “Top industry use cases for real-time analytics”; or to learn more about stream processing in general, and how IBM Streams compares to other stream processing engines, try the eBook “Transform your business insights with streaming analytics”.

IBM Cloud Marketing Manager, Global

More Customer Stories stories
April 30, 2019

Introducing IBM Analytics Engine v1.2 and Announcing the Deprecation of IBM Analytics Engine v1.0

We are excited to inform you about the new version of IBM Analytics Engine v1.2 that will be available starting May 15, 2019. Along with this release, Analytics Engine v1.0 will be retired.

Continue reading

April 16, 2019

Announcing the Deprecation of the Decision Optimization Beta Service

The End of Beta date for the Decision Optimization service is May 17, 2019. The End of Beta Support date is June 20, 2019.

Continue reading

April 2, 2019

Data Refinery and Profiling Changes in Watson Studio and Watson Knowledge Catalog

We'd like to announce data refinery and profiling changes related to Watson Studio and Watson Knowledge Catalog that will take effect on May 17, 2019.

Continue reading