In part 2 of the series focusing on the impact of evolving technology on the telecom industry, we sat down with Vijay Raja, Director of Industry & Solutions Marketing at Cloudera to get his views on how the sector is changing and where it goes next.
Hi Vijay, thank you so much for joining us again. To continue where we left off, as industry players continue to shift toward a more 5G centric network, how is 5G impacting the industry from a data perspective?
5G is really the next big horizon for Communication Service Providers (CSPs) today as it offers them multiple opportunities to move beyond traditional revenue streams and open up new avenues for growth.
Typically, 5G has the potential to offer speeds of between 1-10 gigs per second which is approximately 20x to 30x faster than what 4G technology offers.
On the B2C side, this means faster download speeds, lower latency and that consumers can download ultra-high-definition video on the go. Augmented and virtual reality services are also now possible and will really be brought to life with 5G. Currently, consumers have a wired fiber connection at home too but with 5G it opens up the possibility to provide fixed wireless access as well.
The real opportunity for 5G however is going to be on the B2B side, IoT and mission-critical applications will benefit hugely. 5G can also open up new revenue streams by enabling more connections, and connected devices and businesses can now have up to 1 million connections per km2.
That is a huge game-changer for the industry, for service providers and carriers, considering some of the limitations we have today.
What that means is that this creates new revenue opportunities through IoT case uses and new services. Ultimately, it is not just about speed however, 5G offers enterprise service providers super lower latency and, crucially, more reliability to enable more mission-critical applications. Things like autonomous driving, remote health monitoring or surgeries, smart grids, and factories – all of these require high reliability and low latency to be powered well. This is the next big opportunity for telcos.
From a data perspective, once we have a huge increase in connected devices, there are also opportunities beyond connectivity and companies can diversify into IoT solutions, new platforms, and services. That requires real-time analytics and companies will have to have the capability to ingest and handle real-time streaming data.
AI, edge computing, and cloud are three of the hottest topics in technology today – can you talk us through how they are impacting the telco industry?
All three technologies, AI, edge computing, and cloud, are going to be critical in leveraging the power of the 5G era and IoT to realize new opportunities and to succeed in the industry.
5G and IoT are going to drive an explosion in data. In terms of the IoT, there will be 25 billion connected things and devices by 2025 (Source: GSMA), and these devices will be potentially streaming data every second of the day.
The real question is going to be, how can Telcos capture, analyze, and drive business insights and actions in real-time based on new IoT data? That is where the value of streaming analytics, edge, and cloud is, in how businesses use this real-time data to inform decisions.
A lot of IoT data could end up in the cloud, but not all of this needs to be transported to a centralized location. Much of the data IoT devices generate can be analyzed right at the edge, for example, routine status and health checks for machines including analyzing key digital signatures from sensors such as temperature, pressure, torque reading every second.
Also, data streams that need to be acted upon instantaneously, such as use cases that require ultra-low latency, — things like autonomous driving and hazard detection require instantaneous sub-second response times, and will need to be executed out of the edge itself.
However, the edge cannot function in a vacuum. In order for edge analytics to be successful, you still need the cloud or a centralized data hub, when you can land petabytes of live or test data in the cloud, or a centralized data cluster and then use it to, train, test, and iterate on machine learning models using all that data. These machine learning models, once refined, can then be pushed back to the edge, making it more intelligent – continuously adapting to new and changing circumstances. So specifically for Telcos, and for other industries as well, what we see emerging is a hybrid analytics model, where you will have some intelligence at the edge, but it still feeds a centralized data hub or data lake where you can build more intelligence and predictive capabilities.
And we feel the centralized data analytics in the Cloud or data center will continue to be relevant for three main reasons – one is for context, another is for comparative analysis and finally, there is the machine learning aspect.
Context: Sensor data by itself is meaningless and context is key to IoT. Context allows you to add and integrate to the sensor data with diverse enterprise and legacy data sources for richer context and understanding.
Comparison: Similarly, comparative analysis lets you compare the performance or data across different scenarios or timelines to gain deeper insights. For example – How did the performance of one machine compare to a similar one on the same factory floor? How does it compare to one in another location? How did the equipment performance parameters vary over time? etc.
Machine Learning: And then for machine learning, this requires terabytes or petabytes of test data in order to build, train, and refine models. Once these models are built they can then be deployed back out to the edge in order to make the edge more intelligent.
The benefits of an approach like this are that companies can progressively bring less and less data into the centralized cloud because the edge is smarter and they can make more relevant, more informed, low latency decisions in real-time right at the edge. That is the model that enterprises are now moving towards.
This is why edge intelligence is so critical, but it needs to work in tandem with the cloud in order to be valuable and requires a connected end-to-end platform that manages the entire data lifecycle from the edge to AI.
All three technologies are critical for service providers to really make exciting new IoT use cases possible. To find out more about how Cloudera is supporting Telcos in leveraging the power of 5G, big data and analytics, read our latest eBook on how technology is transforming the industry.