3 Top Trends Driving the Emergence of Operational Analytics
After years of talking about the need for faster data processing, analysis and decision making across industries, the market is finally hitting its stride as the promise of in-transaction operational real-time analytics becomes accessible to organizations of all sizes.
The last six to nine months have seen the emergence of key market drivers creating an optimal environment for the next evolution of the operational analytics economy in 2018, where information is no longer stored away for post-mortem historical analysis, but rather is activated immediately for predictive decision making.
There are three undeniable macro trends driving today’s digital economy that are forcing industries to revamp their data analytics capabilities to stay relevant and competitive in their respective markets.
Internet of Things
The growth of the IoT within almost every industry has enabled more connectivity than ever before. According to industry analysis from Gartner, approximately 8.4 billion IoT devices were in use in 2017 (up 31 percent from 2016), and that number is expected to reach 20.4 billion by 2020.
The rapid expansion of IoT-enabled devices—from cell phones and tablets to industrial equipment and smart meters—is compounding the challenges associated with big data, bringing in more information from more sources at a higher frequency.
Thanks to the IoT, new data is constantly streaming in for use in better decisioning around operations and customer services, leaving organizations with no choice but to figure out how to capitalize on it.
It’s important to note that this influx of information is less about handling the data and more about understanding exactly which pieces of critical information are meaningful—for instance, finding and alerting errors in production, late deliveries, anomalies in processes, fraud in a network, and so on. This is the critical next step to successfully operationalize data analytics in the IoT.
As the IoT expands, increasing the volume and velocity of incoming data, it becomes exponentially difficult for human beings (i.e., data scientists) to keep up with processing, analyzing and ultimately acting on the new information. The solution is often machine-learning technology, which is playing an increasingly important role in driving complex real-time operational analytics to the next level.
Data scientists are already embracing machine learning; identifying patterns and insights from historical information to develop intelligent algorithms that can support business decisions without any human element.
While machine-learning technology is a concept that has been top of mind for a while now, the key differentiator in today’s market is the ability to continuously update the complex logic to ensure the output is consistent with the evolving data patterns associated with new data inputs. For example, fraud and cybersecurity algorithms must be updated regularly to keep pace with the constantly advancing techniques and technologies used by nefarious parties to exploit the new gaps in the ever-expanding sea of data.
The impending 5G revolution represents a turning point that brings all these recent technology trends in IoT, big data, and machine learning to a head. With this next generation network, the scale and speed at which data will move will improve—and increase—significantly.
According to Huawei, maker of mobile phones and networking equipment for enterprise use, while a 4G network can provide thousands of connections for each cell, a 5G network provides up to a million connections per square kilometer, exponentially increasing the number of connections.
With this capacity in mind, the International Telecommunication Union (ITU) is calling for 10 to 20 Gbps (for reference 4G throughput is typically around 5 to 12 Mbps) and a minimum latency of 1 to 5 ms compared to 20 ms for 4G. With the speed of 5G, the promise of edge computing could become reality, streamlining the flow of data traffic from IoT devices and providing in-transaction local data analysis.
Beyond speed, 5G will also bring a new suite of technologies, such as small cells, millimeter waves, multiple-input multiple output (MIMO), beamforming and others—all of which offer a new opportunity for storing and processing data in the cloud, allowing access from a mobile device with effectively no latency. Network slicing, which allows operators to split a single physical network into multiple virtual networks, is another key advancement that comes with 5G.
For instance, a common example of 5G network slicing is the ability to share a given physical network to run IoT, mobile broadband and very low-latency applications (such as cars) simultaneously, even though each of these applications have unique transmission characteristics. The changes that 5G is bringing to the table will enable more advanced data capabilities that consumers will come to expect.
A storm is fast approaching as these technology trends converge at a pivotal point, and organizations that fail to develop a strategy to handle the volume and speed at which data will need to be operationalized will be left behind.