21 July 2016
Every business will soon be an Internet of Things business, according to Cisco Jasper. Theresa Bui, head of enterprise product marketing IoT cloud at the recently merged company, believes it’s “inevitable”, and points out that those who prioritise the development and delivery of value-added connected services will have the advantage.
That’s where you come in. Industry insiders are warning that with the advent of IoT, network managers are required to adopt and learn an ever-growing number of new skills that go significantly beyond those that were required in the pre-IoT era.
For instance, Dava Baumann, VP of global marketing at wireless infrastructure specialist Rajant Corporation, says: “Gone are the days when IT managers’ résumés were based on whether or not they passed their standard network administration certification class and when their primary responsibility was to ensure a high level of network reliability for their WAN, LAN, WLAN, etc. Today, network managers need to acquire specific training across a number of new skill sets that run the gamut, but are primarily founded on a deep understanding of device management, data and predictive analytics, and business processes.”
So why is that? Surely Internet of Things is ‘just another network’ for the IT department to take care of? Few would agree with this. But what is clear is that IoT is about more than just connectivity and has evolved to more of a business type of network.
The new fabric for business
Arqiva launched the UK’s first nationwide IoT network three years ago (see News, December 2014). The communications infrastructure specialist says it is now building two of the country’s biggest end-to-end networks aimed at delivering smart water and smart energy solutions.
Ian Stewart, Arqiva’s director of channel and proposition, IoT, believes that in an enterprise context the Internet of Things should be viewed as being at the very heart of digital transformation: “A smart system which is absolutely integral to the organisation, its business model and its core processes.”
Cisco’s Bui adds to this by saying that what’s special about IoT is that it’s not actually about the ‘things’ – it’s about providing real-time services via those connected devices.
“More specifically, it’s about transforming a business from selling discrete products and services to offering continuously and ubiquitously connected services that can enhance the customer experience. It’s this automated connectivity that is opening up new business models and revenue streams
for enterprises around the world.”
Ericsson agrees. The Swedish company’s name inevitably comes up whenever there is any talk about IoT – the oft quoted prediction of 50 billion devices connected to the internet originates from one of its Mobility reports published a few years ago. Since then, the firm has revised its forecasts. In a recently published mid-year Mobility report, Ericsson now reckons that by 2021 there will be 28 billion connected devices around the world of which 16 billion
will be IoT devices.
Warren Chaisatien, the company’s director of IoT marketing, says IoT will become a “new fabric” of how we live as consumers and how business is done around the world.
“What we’re seeing today is that the culture of IoT has spread very quickly from the backroom of the tech department out into the boardroom of business decision-makers.
“So while the techie guys will play a key role in selecting the technology, the overall strategy and implementation is driven from the business angle. The internet took much longer to reach this status of being at the top of the business decision-maker’s agenda. The agenda has now been balanced between being technology-driven and business-driven.”
As a result, many of the experts we spoke to support the overall view that IT doesn’t necessarily have sole responsibility for a corporation’s IoT.
For instance, Paul Lipman, CEO of online security specialist BullGuard, says: “On the surface you would think that IoT is owned by the IT department. However, the real owners of IoT are far more likely to be within the business. The rationale is simple. Most IoT equipment within the enterprise is designed to gather data, whether it’s to control and manage machines, identify potential equipment, or gather data on customer actions. As a result, we’re likely to see IoT actually owned by departments within the business who use the information generated by IoT devices.”
As an example, Lipman says this could be a product development department that pushes out new features and functionality to customers on a regular basis, or a marketing department that uses IoT data to lock-in customers with personalised services and products. Of course, the IT department will underpin all this, and will be required to roll out devices and manage the network in order to provide the foundation.
Rajant’s Baumann concurs: “At the end of the day, the IoT must be owned by IT and integrated throughout the enterprise. This is based on the idea that if IoT is properly integrated, organisations will be better positioned to leverage large amounts of data from which they can
act upon and improve decision making.
“For example, if SCADA (supervisory control and data acquisition) data can move from a centralised function and integrate with other decentralised data points, the collective data and subsequent analysis of them will likely offer more intelligence and insight for operations. If the data cannot be integrated with other sources of data and acted upon, it’s just another standalone network with a data silo.”
OT versus IT
So the real IT skills enterprises are looking for when it comes to IoT is arguably not network management but data management.
In Arqiva’s view, the network itself is only one component of IoT, along with the sensors embedded in devices and the software to extract and analyse the huge volumes of data that come out of them. And even here you have to be careful, according to Stewart: “IoT is not all about moving large volumes of data in real-time. It’s more about getting the right data and doing the right thing with it. In reality, the value of the data from a smart system is about giving the context to a particular situation, which then enables the user to take the appropriate action or make the best decision.”
Cisco Jasper’s Bui says today’s conversations about IoT are being had within two different camps: OT (operational technology) and IT. She reckons the struggle is how they can blend their IoT efforts with the rest of the organisation’s IT environment.
“IoT ownership is still very much siloed, but as IoT goes mainstream we’re seeing IoT data highlight the need for these organisational challenges to be resolved. Many organisations are still adapting to this new paradigm.”
Bui reckons that in the next five years or so, the industry will see the creation of the role of a chief data officer to deal with these challenges. She says they will oversee the integration of IoT and IT systems to enable the entire organisation to access IoT data but leverage it for different purposes. “OT is already starting to use the same set of tools that IT is using and this means the entire company will be able to see the full picture and become aware of the impact of the connected services they are delivering.”
But in order to do that, those responsible for the IoT in an organisation will first need to shift away from a mindset that is product-oriented to one that is geared towards service. And that, says Bui, leads to the first challenge and skill that network managers will need: cloud proficiency. “Businesses need visibility and real-time control to manage connected services, and ensure the best possible experience for their customers. SaaS models are ideal for these IoT businesses.”
Many vendors such as Ericsson consider scalability to be one of the top IoT challenges for network managers. “At one level they are connecting humans but of course going forward they are connecting a lot of things,” says Chaisatien. “So suddenly there will be massive growth in the number of devices that need to be connected.”
According to Bui, this is where automation becomes vital. She says that even if a company deploys a “modest” number of devices – say 100,000 – each one will still need mobile service management, real-time engagement, support diagnostics, billing, provisioning, etc. “To fully optimise performance and cost, an average device requires 100+ real-time actions per year. On a 100,000-unit deployment, that means 10 million actions every year. At that scale, manual processes are inefficient, costly and simply impractical. Automate every process possible to reduce response times and streamline operations.”
How safe are your things?
While the network manager’s traditional duties of maintenance and management also apply to IoT, BullGuard’s Lipman says there will be greater emphasis on security: “With IoT, the network manager’s role could extend in a number of directions but one central aspect that will need to
be taken on board is security.”
US-based managed network services provider KORE Wireless agrees. Its CTO Gavin Carter says that while many of the principles of running an IoT network are the same as running a conventional LAN or WAN, it’s all too easy for those involved to lose discipline and forego security concerns.
“The variety and availability of connected devices can create a false sense of maturity, particularly when it comes to security. So I would say the most important skill for the network manager is to focus on security.
“Perhaps the most famous example of a lack of focus on security while driving an IoT strategy is US superstore Target where refrigerators were connected to the same network as the payment systems. Their desire to take advantage of the IoT outweighed pretty basic security considerations.”
Many commentators therefore believe IT departments therefore need a new level of competence as they secure not only human controlled devices on networks but also the massive number of ‘things’.
“Security is key because in the world of IoT every connected device is potentially a vulnerable point for your organisation to be exposed to,” warns Chaisatien. “The first thing that many organisations do with IoT is connect devices and things. But all that’s just a starting point. What do you do with the data to turn them into insight? And in many cases, if you go up the IoT value stack, the data goes up to the application level. So security has to work from the device level and to the network level, back into the cloud, data centre, and then in the software level in the application at the end.”
Lipman echoes all of this. He says the sprawling nature of IoT requires comprehensive management of the entire network right to the edge and the network entry point, as devices seek access and data transfer to core network components. But he says this won’t be straightforward. “For instance, at the traditional switch level, connected devices and sensors will be transmitting data on the network. But given the potential scale of IoT networks and the quantities of data that are being collected, sending data from devices straight to the data centre could have an impact on performance.
“Furthermore, simply monitoring and controlling the flow of data packets to and from IoT devices will not be enough to guarantee security. All devices right out to the entry point into core networks must be made smarter by network management and the switches on the network.”
Lipman goes on to point out that not every ‘smart’ IoT device is actually smart, and within the security context some are actually “decidedly dumb”. What he means is that security is rudimentary and easily breached with these devices, and that represents a threat to network security.
“As a result, there will be a great onus on network managers to ensure that software or firmware patching functionality is central to any IoT device. At the same time, they will need to also ensure that IoT devices have system resets and unique passwords, and that there are no back doors that can be exploited.”
Powering the WAN
Arqiva’s Stewart points out that depending on the application, some traditional connectivity technologies may not be appropriate. He says that while GSM and Wi-Fi, for example, can handle higher bandwidth demands, much of the data generated from IoT doesn’t necessarily need high bandwidth computing power. This is where low-power wide area networks (LPWANs) come in.
France-based Actility provides a set of tools to deploy, manage, operate and monetise IoT and M2M networks based on LoRa (long range) WAN technology. According to Gabor Pop, the company’s solutions marketing manager, network managers have to understand that deploying an LPWA network such as LoRa is “not that difficult”. He explains that usually a couple of base stations are enough to provide connectivity over a very long distance. And to enhance in-building connectivity, nano- or picocell gateways are available to provide coverage in areas where signal propagation is complicated. “A network server that can be installed on premise or in SaaS mode is then required to manage the network communications along with an OSS to manage device activation, security and RF issues. The required skills can be mastered in a couple of days.”
Pop says the main constraints for IoT are currently battery life and communication range. “Today, Wi-Fi is clearly enough for short range, high data rate and powered communication – but it consumes a lot of battery. 3G/4G is great for long range communication – but you still have to recharge your phone every night. Bluetooth Low Energy or Z-Wave are great for short range communication – but you need a local router/receiver.”
He reckons an IoT WAN technology such as LoRa provides the solution as it allows communication up to 15km with battery life of several years. “This means that you can get rid of any local gateway or repeater, and you can forget about your IoT devices as you don’t have to recharge them. Very much like in LTE, dedicated networks for LoRa allow for objects to communicate directly through public networks.”
Patented by Semtech and the LoRa Alliance, LoRaWAN is not the only specification developed for battery-operated IoT devices in LPWANs. Some of the other platforms include: Narrowband IoT (3GPP’s standard for cellular networks); NWave and Weightless; Sigfox (as used by Arqiva); and others. While various standards bodies, such as oneM2M, ETSI (see News, p2), et al, are working to develop a universally accepted interoperable standard for connecting IoT devices, this does not as yet exist.
Many of the vendors we spoke to did not seem overly concerned by this. They say the ‘fight for standards’ is typical with all new technologies such as IoT which is still in its nascent stages. Ericsson’s Chaisatien believes the pace of development with IoT is fast and so the market will soon settle on standards.
In the meantime, he advises network managers to look for standardised technologies.
“Even Ericsson, which is a large ICT provider, cannot provide everything. We have to put bits and pieces together. So we have to partner at the device and chipset level with Intel and Qualcomm, for example, and those companies all use standardised technologies. The best advice for enterprises and decision-makers to follow is to look for the companies that have the most traction.”
From a technology point of view, the network is clearly an essential building IoT block. But like Chaisatien, Stewart believes the other critical success factor to unlocking IoT’s potential is understanding the whole ecosystem that surrounds it – the specialist sensor and device providers, the data analytics applications, the software platforms, etc.
“It is the role of the network manager to select and manage the right types of network to meet the needs of the organisation. However, if we take the full end-to-end view of the IoT, then that role will also go way beyond keeping the lights on.”
Stewart reiterates that network managers to think of IoT as a key enabler for the digital transformation of their organisations and business models. “It’s not the volume of the data coming out of the IoT. It’s what you do with it that counts. Use it to drive better customer insights or enhance the customer experience. Use it to improve core operational processes and inform decision-making. Use it to create new products or services, or improve existing ones. Take a wide-angled view.”