Networks are being significantly impacted by the commoditisation of hardware and the migration of functionality and value to software. This dramatically increases network complexity, but more importantly gives carriers an opportunity to gain a competitive advantage through disruptive new services.
One lucrative new market that can be accessed through more innovative use of software is Industrial IoT (IIoT). For example, Statista recently found discrete manufacturing, transportation/ logistics, and utilities industries are projected to spend $40 billion each on IoT platforms, systems, and services by 2020.
There are many software technologies that promise to revolutionize networks, including network slicing, ONAP and software-based architecture. However, it is that AI offers a transformative opportunity through its ability to industrialise large scale pattern recognition and automate complex processes.
AI-augmented core and Radio Access Networks (RANs), for example, are set to offer carriers an unprecedented level of control and flexibility to develop higher performance services, providing a clear route to access mission critical applications such as IIoT, automotive and healthcare.
AI in the RAN
Until now, optimising the RAN has focussed on improving ‘best effort’ accessibility, with AI addressing Self-Organizing Network (SON) activities, such as load balancing and neighbour cell relation management.
In a way that has never been possible before, AI can enable carriers to rapidly analyse vast RAN and device datasets to optimise networks dynamically in line with customer SLAs. This is set to transform RAN performance and reliability, both in geographically focused private networks and in public networks where the existing deterministic approach does not scale economically.
To meet the Quality of Service (QoS) demands of IIoT and to overcome their challenging RF environments, it is necessary to optimise each industrial network for a range of criteria. This level of customisation must consider the specific coverage and propagation requirements of individual deployments, guaranteeing low interference and error free communication. This can be achieved efficiently through AI, which will significantly reduce latency by removing High Availability Resolution Queue (HARQ) and error correction mechanisms.
Benefits from AI in the core
In the core and cloud, AI techniques will drive orchestration, optimisation, service discovery and optimum path configurations. Hence, the automation AI brings will allow increased network complexity to become more manageable.
This will allow carriers to offer new and highly targeted services to enterprise customers, moving away from commoditised infrastructure and ‘best effort’ QoS. As a result, carriers will be able to offer densification and increased levels of network control, allowing individual industrial components to be managed carefully. Using this approach, carriers can enable IIoT platforms to extend beyond their existing wire connected fixed machinery and into mobile equipment, such as robots, augmented reality systems and distributed sensors.
By provisioning time critical activities in the device, or at the edge, and locating best effort compute activities in lower cost, centralised cloud infrastructure, AI techniques can provide agile, task optimised implementations. AI can also be leveraged in areas such as customer care and service management where Deep Learning technologies can anticipate end-user requirements or trigger preventative maintenance.
The critical building blocks to achieve these gains include high performance wireless or wired networks, hybrid cloud implementations, an AI application platform plus AI network operation and optimisation.
AI adoption in telecoms
Momentum is building rapidly in telecommunications, where early adopters are already realising the benefits of AI in customer service. For example, Forbes recently reported that AT&T use chatbots to handle online interactions, Comcast use Talking Guides to provide help via voice-recognition, whilst CenturyLink use systems to interpret and respond to inbound emails with staggering accuracy.
From its dedicated AI research labs, Cambridge Consultants works with clients across the wireless ecosystem, helping them capitalise on novel AI techniques to create unique commercial propositions ahead of the market.
As discussed in our recent report, AI: understanding and harnessing the potential, the advent of a small number of transformational technologies is going to have significant impact on carriers and their supply chain.
The impact on enterprise and industry could be as significant as the change commerce went through when the Internet first became widely adopted during the 1990s. Just as it is unthinkable for a business not to have omnichannel customer engagement today, it will soon be unthinkable that customer relationships will end with delivery and payment. Whether implemented in the device or through large scale networks, AI will revolutionise all industries and accelerate the trend for customer-vendor relationships that last the entire product lifecycle.
Cambridge Consultants’ AI experts will be discussing AI in more depth in RCR Wireless’s upcoming webinar “How AI will transform telecoms” on 18 October. We hope you can join us then.
The post Why AI is essential to success in Industrial IoT appeared first on RCR Wireless News.