Your phone: Smarter in 2017, says Deloitte

On-board machine learning capability could be seen in handsets

Deloitte, in its annual Technology, Media and Telecommunications Predictions, is tipping smartphones to get on-board machine learning capability in 2017. Other predictions include a sharp increase in the size and frequency of DDoS attacks and rapid growth in IT-as-a-service.

Deloitte predicts that over one fifth of smartphones sold worldwide in 2017, around 300 million, will have on-board neural network machine learning capability.

“They will allow smartphones to perform machine learning tasks even when not connected to a network,” Deloitte says. "This functionality will enhance applications including indoor navigation, image classification, augmented reality, speech recognition and language translation.”

Stuart Johnston, Deloitte’s national TMT leader for Australia, said: “Many of these functions already exist today, but through a connection with the network. The phones being released will be able to do this without the network, and that will be quite a significant advance.

“It will make the mobile phone have a greater impact in our lives. It will start to learn about our behaviours through things like pressure sensors and start to predict our health. It will become more and more integral to our live, and better able to interact with surrounding objects.”

He added that this development would be particularly significant in Australia, one of the world’s leading adopters of smartphones, and he suggested that smartphones might be further along this track than many people realised.

“We are already using this functionality even though we not know it. … Some of this functionality is coming from iOS and Android updates, and from developers. Sometimes we won’t even know it is there.”

Indoor navigation on the rise

Related to this development, Johnston said, would be the increased availability of indoor navigation on smartphones. Deloitte says indoor precision digital navigation presently accounts for less than five percent of usage, but is tipping this to rise to at least 25 per cent by 2022.

“Outdoor navigation has been calculated to have 0.4 per cent positive impact on US GDP, and it has been estimated we spend 90 per cent of our time indoors, where GPS does not work,” Johnston said.

He told Computerworld: “The winners will be those organisations that really think about how to make the users’ experience better rather than just trying to push more ads. There are companies that are doing that really well, and they are the ones that are winning at the moment.

“There are a number of companies in Australia that are well positioned the question is whether they can improve their position rather than defending the status quo.”

He added: “I don't think it will be very apparent to consumers this year: there will be a lot of trials and experiments, but Australian retailers don't step in, it will come in from somewhere else.

Security with one finger

Another smartphone development tipped by Deloitte is the growing use of fingerprint recognition: it estimates that by year end there will be close to one billion smartphones and tablets equipped with this function in use around the world. In developed markets they will account for about 40 percent of the smartphone population and the functionality will be used regularly by about 80 percent of users.

Stuart Scotis, a partner in Deloitte’s TMT practice, said: “Fingerprint recognition is absolutely mainstream… There is now an opportunity to integrate fingerprint recognition into more and more applications: retail, financial and even for building security.”

He added: “The fingerprint does not have to be stored in the cloud to enable access to your phone, so there is minimal risk of it being exposed online, and that changes the game in the way we manage biosecurity.”

More and bigger DDoS

On DDoS, Deloitte is forecasting at least one 1Tbps DDoS attack every month in 2017, after 2016 saw the first two attacks above this threshold. It expects the total number of DDoS attacks to pass the 10 million mark, up from 8.4 million in 2016, and to reach 17.4 million attacks in 2020.

Deloitte attributed growth to the increased number of insecure IoT devices, availability of online instructions for launching DDoS attacks and a general increase in uplink speeds.

Stuart Scotis, a partner in Deloitte’s TMT practice, said that, to mitigate these types of attacks, organisations would need to become more sophisticated.

“They will need a multi-dimensional strategy and any defence that is predictable, such as rerouting traffic, will need to be dynamic in nature because the attackers are getting more sophisticated and are starting to anticipate some of the responses to a denial of service attack. So organisations will increasingly need multidimensional approaches to defend against DDoS attacks.”

Join the Computerworld newsletter!

Error: Please check your email address.

Tags Deloittesmartphonesmachine learningartificial intelligence

More about DeloitteTechnology

Show Comments

Market Place