Following the purchase of Alcatel-Lucent in 2016, Nokia is aiming to become a prominent force in the internet of things (IoT). Jason Collins, vice-president of IoT, applications and analytics at Nokia, talks to Abi Millar about navigating the most interconnected era in history.
In November 2016, the telecoms giant Nokia finalised its purchase of Alcatel- Lucent, which was merged into its Nokia Networks division. The idea was to improve connectivity, expand its product portfolio and, most importantly, take a flying leap into the internet of things (IoT).
“During the past three years, Nokia has reinvented itself, emerging as a leader in the technologies connecting people and things. We are focused on playing a central role in a world where everyone and everything will be connected,” said president and CEO Rajeev Suri following the purchase.
The acquisition was certainly timely. With 5G just around the corner, all telecom providers are trying to work out how they can shake up their offerings, developing the kind of highperformance networks best suited to a hyperconnected world.
Nokia is no exception. In 2016, the company demonstrated the world’s first 5G-ready network, and it’s already deploying ‘4.9G’ technology in an attempt to bridge the gap with what’s to come. Rebranding itself as a full-scale IoT vendor, it sells network infrastructure, devices and sensors, all under the banner of its IoT platform, Intelligent Management Platform for All Connected Things (IMPACT).
“There are a number of areas where we think we can add value,” says Jason Collins, vice-president of IoT, applications and analytics at Nokia. “Our traditional business is focused on the evolution from 4G to 5G, so we will offer a portfolio of radio and networking gear optimised for IoT applications. The area I represent is the layer above the networking – the applications and analytics software business.”
Collins has a lot of experience in this area, having previously headed up IoT ecosystems and business model development at Alcatel-Lucent. At Nokia, he focuses on two types of value propositions – one broad in scope, the other more bespoke.
“The first is a horizontal infrastructure layer, where we focus on building the relationship between devices and applications,” he says. “Every application has the same fundamental needs and a good chunk of the problems could be genericised. The second has to do with building end -to-end vertical solutions in domains where we think we have a strong play.”
The first proposition, IMPACT, is an application-independent platform suited to communication service providers, enterprises and government. It is capable of supporting 80,000 different device types and can handle everything from data collection to application enablement. Regarding the second proposition, Nokia is particularly interested in working with industries such as transportation, utilities and healthcare. “There are a lot of industries that can use this infrastructure, so you will see Nokia making key select choices in areas where we can, and will, excel,” explains Collins.
City of the future
Probably the most talked about use is the emerging concept of the ‘smart city’. In essence, a smart city takes data from various physical devices and uses that data to optimise the way it runs. Nokia is working directly with a number of cities to address the problems they’re facing.
“As cities grow, that puts a lot of stress on infrastructure and anything that is limited in supply, whether that’s energy or parking spots or public transport or the roads themselves,” says Collins. “On top of that, cities in their traditional form are not managed in the most efficient way. Typically, they are broken down into departments that aren’t allowed to share investments or data, even though it would be more efficient if they worked together.”
He believes that a truly smart city would move away from this siloed approach and take a more holistic overview of its needs. Ultimately, it would look towards an architecture that could meet several of those needs in tandem. If I’m connecting my lights and my door lock to an iPhone app, that’s great, I can control stuff remotely, but how do I know I can trust the remote control entity to be authentic?
“Cities need to measure movement of crowds, monitor environmental noise, conduct law enforcement for public safety and so on,” he says. “A lot of these things can be done with a common infrastructure. One ideal point for deploying radios and networks in the city is streetlights – there are streetlights everywhere, they’re interconnected and they happen to have power.”
A smart streetlight might function not only as an LED, but also as a 5G access point, environmental sensor and crowd-monitoring video source. All these functions could reside in the same assembly on the light post, with multiple stakeholders in the back office consuming the data.
With this example in mind, it is easy to see the advantage of Nokia’s platform approach. You might only need one solution now, but that solution may not be fully equipped for the sort of demands you’ll face in future.
“Some customers might say they just need a parking solution today,” says Collins. “Our response is, you can buy a parking solution from multiple vendors, but with a platform approach, your solution is ready to expand into environmental sensing, waste management and so on. We try to coach customers to not only look at the immediate use, but also their future needs and the total cost structure over the life cycle of the project.”
Safety on the roads
Aside from smart cities, another much-discussed IoT application is driverless cars. Here, the idea is that AI technology could be used to make the roads safer, eliminating the deadly costs of human error.
While Nokia is not an automotive manufacturer, it does offer plenty of relevant technology. “To give one example, driverless cars use an AI algorithm that needs to be trained how to behave in certain traffic conditions,” Collins says. “The amount of data it takes to generate a reliable algorithm for self-driving is massive – you need tens or hundreds of thousands of cars with data probes to create reliable data sets. We can build very secure and efficient communication infrastructure between those cars and networks.”
He adds that, while fully driverless cars may still be some way off – more for regulatory than technological reasons – their eventual deployment is inevitable. “If you look at statistics on how many people die every year and how much money is lost in accidents, you’re talking about an industry that’s well over a billion dollars or euros a day,” he says. “Driverless cars will happen no matter what, because there’s too much money involved in avoiding accidents.”
Of course, as the IoT revolution gathers pace, it will pose challenges. Aside from the wave of job losses likely to accompany automation, connectivity on such a massive scale will create security concerns. To put it simply, as soon as a device gets connected, it becomes accessible to undesirable entities that want to take control or steal its data. We only need think of the recent WannaCry ransomware attack, which hit more than 200,000 computers in 150 countries. It is hard to imagine the kind of chaos that would have ensued if IoT devices had been the target.
Who takes control?
Collins feels there are two key ways of managing this risk; the first being a matter of trust building. In essence, developers need to create a two-way authentication model to ensure the users are who they say they are.
“You cannot trust data from an unmanaged source – it sounds very simple, but it’s so overlooked in the industry,” he states. “If I’m connecting my lights and my door lock to an iPhone app, that’s great, I can control stuff remotely, but how do I know I can trust the remote control entity to be authentic? If you don’t think carefully about that question, you quickly get yourself in trouble.”
In addition, developers should make sure they understand their device’s expected patterns of use. This means that when anomalies arise, it’s easy to flag them up and identify the culprit, in a similar fashion to how a credit card might decline an anomalous transaction.
“If you connect a parking meter to the internet, it’s very predictable what kind of conversation the parking meter should have with the outside world – you can start to do machine learning on what’s a normal pattern of behaviour,” Collins explains. “If some entity from a foreign country logs in and finds a way to access the data, a good security design will alert you to that and say there’s something abnormal happening on the meter.”
He believes there is a lot to be done in creating industry standards and best practices for data management. Currently, such standards exist in specific verticals, such as the HIPAA privacy regulations within healthcare, but in the future they will need to become increasingly far-reaching.
Naturally, these issues have ramifications stretching beyond IT departments. With the number of connected devices expected to surpass 20 billion by 2020, questions about data security and industrial automation are going to become front and centre in every sector, and C-suite executives will need to be attuned to customers’ concerns.
“What’s important is you’re very transparent and up front in what you’re going to do with people’s data and what your intentions are,” concludes Collins. “There’s a lot of concern, and probably rightfully so, about how personal data will get used or abused. So you’ve got to be very trustworthy if you expect people to subscribe to your vision and give you access to their data, which I would consider to be the new gold standard.”