As AI technology continues to develop, more and more industries are beginning to dip their toes into artificial waters. 2018 is set to be a huge year for machine learning, neural networks and artificial intelligence, and the telecoms industry is investing heavily in these technologies to improve their services.
After all, it’s an unfortunate truth that the telecoms industry received the second most complaints in 2016, just behind retail; a reflection on the growing pressures placed onto internet and mobile providers to deliver a near-perfect service, across the entire UK.
AI technology could revolutionise the way telecoms businesses operate – let’s take a look at how the technology is currently being applied, and the impact it will have.
AI chips in smartphones is one of the most recent strides in telecoms technology. The iPhone X’s A11 Bionic chip and Huawei’s Kiri 970 chip contain a Neural Processing Unit (NPU), which takes pressure off the Central Processing Unit (CPU) by freeing it up to perform other tasks, helping the phone to run faster (and multitask). AI chips process machine learning or deep learning tasks, such as image recognition.
Huawei says its NPU can recognise 2,000 images every second. Apple’s AI chip, meanwhile, focuses on face and voice recognition, recording ‘Animojis’ and working out what you’re trying to frame when taking a photo; it can handle an incredible 600 billion operations per second. It’s clear that as AI becomes a bigger part of our lives, we’ll see competitor smartphones following in the footsteps of the main players, and adding their own AI chips into their latest devices.
The European Telecommunications Standards Institute (ETSI) has established a group to investigate how AI could be used to improve the way that networks operate. Currently, it is focusing on autonomous, self-managing networks, which would be able to automatically adjust services based on several factors, from environmental conditions, to user requirements and business goals. This system would be able to ‘learn’ from experience, allowing it to configure networks to ensure demands are met, which will ultimately improve network use and reduce costs.
It’s unclear how far away such technology is from being implemented in the real world, but it could revolutionise the way that services like broadband and phone networks are delivered. The system – once experienced - could operate without human intervention, and would always ensure it is keeping up with consumer demand.
One area of the telecoms industry in which AI is already making a huge impact is customer service. Broadband, mobile and landline providers are inundated with customer queries, and it’s important to ensure that every customer feels valued, and can get their issue or query sorted with minimal wait times.
Vodafone’s TOBi is the first live chatbot in UK telecoms, and it’s made a big difference to the way the company’s customer service team operates. The chatbot currently resolves more than 70% of customer queries without human intervention, and it is constantly learning, so it can deliver a better service. Of course, an AI can’t fix every customer query, which is why TOBi is programmed to recognise when a real human is required and seamlessly hand the customer to an operator.
Eventually, Vodafone plans to fully integrate TOBi into its systems, which will allow the majority of webchat conversations to be automated, freeing up human employees to carry out tasks an AI is incapable of resolving.
Telefonica, the parent company of O2, is taking a different direction with AI. Last year, it announced the creation of Aura, a voice recognition software which can assist customers over the phone, rather than over the internet. Set to launch next year, Telefonica says Aura will help the company to better understand their customers, boosting loyalty and trust.
Total Access Communication, Thailand’s third largest mobile phone provider, is also using machine learning and AI to help improve its customer service. AI-powered chatbots are currently being tested, and calls are being analysed using big data analytics, so that the company can understand the queries most commonly made by customers. The company is also using machine learning to determine which call package an individual customer needs, based on their usage. This application will only get better over time, and could be a real asset to the company.
Bad press may have plagued telecoms businesses in the past, but it’s clear the industry is investing big in technology that ensures a better service and product for all. For now, it will be interesting to see just how big an impact AI continues to make on telecoms, and whether customer sentiment improves as a result.
How can we help?
How can we help?