Forecasts and estimation of the number of IoT devices deployed globally continue to rise and rise again. Few industry analysts or researchers agree on what that number eventually might be but all accept that there will likely be upwards of 30 billion of them by 2030 and that, by that time, the market will be worth US$ 20 trillion or more. Such massive figures would have been dismissed as laughable when the notion of the Internet of Things first filtered into public consciousness back in 2010. However, rather than undergoing the sort of steady linear expansion that characterized the "one device, one user" epoch (i.e. one mobile handset per subscriber or one PC per owner) the growth of IoT has been explosive and close to the point of becoming exponential. The fact is that there are already many millions of IoT devices and sensors already out there - and more and more are coming online every day.
The remarkable direct result of the mass deployment of IoT-connected devices is that the internet we know and use today is very much bigger and more expansive than it would ever, or could ever, have been without them. Furthermore, the social and economic potential engendered by the proliferation of IoT is both immense and, as yet, barely tapped. Consider the way the system works: connected devices manufactured in, say, the US, are sold to customers elsewhere around the world where connectivity is provided by telcos and network operators. Those devices are then tracked to and used in many other countries and thus the global wheel of IoT revolves and increases pace.
However, although the potential is evident and acknowledged, many CSPs have found it hard to generate revenues from IoT traffic. That's because detecting, measuring, classifying and monetizing IoT is proving difficult because service providers have to be equipped to be able automatically, accurately and quickly to provide appropriate service levels at the right price for each and every IoT segment and use case, and there are so very many of them. The fact is that pressure on network systems to detect and classify devices via signaling and usage software and analytics is tricky and places a heavy demand on all-too finite and increasingly stretched resources. However, there is a prime solution to what is a particularly difficult set of problems: Artificial Intelligence (AI), automation and 5G.
Solving global IoT roaming
With an AI-powered 5G network, telcos, CSPs, DSPs and network operators will be the loci of a communications revolution, able, in real-time, to connect, manage, monitor and monetize a huge array of personal devices, plus autonomous vehicles, robots, cars, cameras, domestic equipment, sensors and much more. The growth in IoT roaming traffic is accelerating, spurred by the effects of the global spread of Covid-19 which revealed the extent of devices already in permanent roaming and the increasing reliance on them by the enterprise and the individual.
Applied AI will solve the problem of global IoT roaming not only by segmenting consumer and IoT traffic but also via leveraging signaling and usage analytics to provide insights on roaming traffic patterns, (such as permanent roaming and silent IoT devices) and uncovering and isolating abnormal activity. When used in combination with the power of 5G, CSPs will have new business opportunities by providing service differentiation to different IoT segments such as massive, critical IoT or specific enterprise, which will allow them to move up the value chain.
How then will the coiled potential of IoT be sprung? What capabilities will telcos need to have in place to be able to take advantage of the opportunities presented and benefit from them? IoT collects massive amounts of data from countless multiple constant sources but to read the patterns and then fully exploit the power and relevance contained within all that disparate data they must quickly be collected, processed, managed, analyzed and acted upon to unlock accurate meaning. Only then can meaningful insights be gained and the context and creativity be provided to make informed decisions and drive smart actions in fast and agile responses to dynamic situations. The system must also be able to balance the requirements for localized and centralized intelligence, and personalization with full regard for, and compliance with, data protection legislation, confidentiality and data privacy.
Changing the world of telecoms
When used in tandem with machine learning solutions AI can also accurately measure and predict operational conditions and detect the parameters that need to be modified to ensure ideal outcomes. Intelligent IoT can analyze and describe processes such as those that are taking too long, using too many resources, are failing or redundant. What's more, they can also determine exactly which tasks and routines need to be fine-tuned to increase efficiency. The net result will be quantifiable benefits accruing to enterprises and consumers via intelligent automation, proactive intervention and personalized service and experience whilst AI's predictive capabilities will help to eliminate bottlenecks and network downtime.
Whilst the number of mobile devices and even high-end computers integrated with IoT are continuing to multiply, the most usual IoT ecosystem continues to provide connectivity to a huge number of small, inexpensive low-end sensors that either continually or periodically transmit tremendous amounts of signaling data that can put strain on, and occasionally totally overwhelm, the network. However, an AI-enabled IoT ecosystem provided greatly enhanced scalability by both analyzing and summarising data derived from one particular, uniquely identified device. In this way, AI enables massive data volumes greatly to be simplified whilst incoming data to remain completely meaningful.
The top five benefits of the integration of AI with IoT to help transform a business and maximize the value of a network are these: Better and deeper relationships with customers, Cost-effectiveness, Increased operational efficiency, Highly secure and safe network environments and ecosystems, Ability to focus on new products and services. Thus IoT has the potential to change the world for telecoms operators and their client bases of enterprises and individual consumers. As the sector evolves further, the exploitation of the flexibility of 5G, and the additional power of machine learning will enable telcos to leverage the myriad new data insights they gain to benefit their customers and themselves.
Jordi Castellvi, Product Management Director, TOMIA