IoT set to push computing to the edge in 2018

The internet of things has steadily grown in importance over the last few years and because of this we thought it was worth publishing some specific trends on the subject for 2018. To compile these, we spoke to a number of individuals in the field, and the following piece is based on their feedback, along with our observations on the industry from the last 12 months.

We have also included bespoke trends—from Cisco in the networking space and Thales eSecurity in the security space—which can be found at the bottom of this article.

More connected devices

The number of connected devices will continue to grow and most of the analyst houses are already weighing in with specific predictions. This connectivity has gradually infiltrated our homes and workplaces and is slowly promising to be a greater enabler for businesses. “As discussions about IoT begin to filter up to the boardroom, they will increasingly become part of the strategic agenda,” suggests Andrew Till, VP for technology, partnerships and new solutions at Harman Connected.

A greater emphasis on security

The security issues associated with IoT are not set to away anytime soon. A recent Kaspersky event highlighted this very well in the consumer space with an hour-long demonstration on how to secure the home network with a homemade Raspberry Pi solution. This was necessary because there is still nothing off-the-shelf available.

Things are a little different for businesses, of course, but not as much as you might hope. As Sanjay Srivastava, chief digital officer at Genpact points out: “Manufacturers will continue to make the mistake of not introducing security into the design phase.”

The growing importance of analytics

The internet of things covers a number of different areas—from hardware to connectivity—but the most important aspect for many businesses is the data generated. The trouble is once this information has been captured it needs to be processed and this is where other important trends, like machine learning, can be utilised to improve its value. “The focus for IoT will shift from connectivity and onboarding to more on analytics and machine learning,” suggests Vijay Raja, IoT solutions lead at Cloudera.

“Advanced analytics and machine learning will start to play crucial role in IoT, as organisations will increasingly focus on driving automation and intelligence into their operations, based on all of the data that is generated by connected devices,” he adds.

The need to process data at the edge

The strategic growth and increasing volume of data has, in turn, has made the ability to process data close to the source—instead of pushing of pushing it straight out to the cloud—very important for many businesses. “With data proliferation coming from the rise of IoT and the predicted capabilities of 5G in 2018, edge computing will become ever more vital,” says Neil Bramley, B2B client solutions business unit director of Toshiba in northern Europe.

“For organisations that handle large amounts of data, deciphering what to send to the cloud can reduce backlogs allowing it to perform the heavier tasks whilst edge computing technology allows increased mobility and real-time processing thus increasing efficiency at both ends of an organisation’s IT chain.”

Blockchain could help deliver on the IoT promise

It is possible blockchain could help secure IoT and deliver on its promise. IOTA, for example, launched a bespoke (and as yet unproven) blockchain for the Internet of Things recently. “In the world of IoT you’re generating millions of small transactions that are being collected from a distributed set of sensors. It’s not feasible to operate these systems using a centralised transactional model: it’s too slow, expensive, and exclusive,” explains Ettienne Reinecke Dimension Data’s group CTO.

“To extract the true value from IoT technology you have to be able to operate in real time. Once a sensor alert is received from a control system you must react to it, meter it, and bill for it instantly – all of which negates the viability of a centralised transactional authority. The cost of the transaction has to be near-zero or free, and the cost elements of a centralised model simply don’t support the potential business model in IoT,” he adds.

IoT trends: A network perspective from Cisco

These predictions come from Theresa Bui Revon, director of strategy

1. There will be a battle over IoT data ownership: In light of highly publicised security breaches throughout 2017, everyone is concerned about their data. And the IoT is generating more data than ever before. In 2017, IoT and Big Data intersected in a big way to collect, analyse and make sense of IoT data. In 2018 the industry will move forward into a new era – one that requires figuring out who gets access to that data, how can it be used, and who owns it. And with only 1% of IoT data actually being used today—it is now critical to determine which data is actually valuable and actionable to help drive real results.
2. Artificial intelligence and machine learning will help secure IoT: IoT security incidents continue to increase. Rather than reacting to incidents after they happen, 2018 will be the year the industry takes a more proactive approach – thanks to artificial intelligence (AI) and machine learning. Combining IoT data with AI and machine learning technologies will enable the industry to predict—and thus prevent—attacks before they happen.
3. IoT interoperability finally becomes a reality: Devices made by different manufacturers are finally integrating to work together, and device interoperability will become mainstream in 2018. This is finally happening because of the adoption of standards like OMA Lightweight, which is being adopted at the chip, software, and hardware levels so that these devices can speak a common language (or send and receive common data to each other). This drives an even greater need for IoT platforms that can apply policies and automate what actions are taken based on the data being shared.
4. Low power wide area networks (LPWANs) will make IoT more accessible, but they need a business model: In 2018, the availability of LPWANs will make the connectivity of devices far more affordable and will increase adoption, but this will create urgency around the need for a business model to make LPWANs profitable. Because low power devices typically use very little data, you can no longer charge based on the amount of data used per device. Instead, LPWANs will need to charge based on the business outcomes delivered, not the data used in delivering those results.
5. Transportation as a Service (TaaS) goes mainstream: By the end of 2018, most major automotive companies will commit to a percentage of their fleets being used as part of a ride-sharing service. Some will invest in existing ride-sharing services, while others will introduce their own branded offerings. Likewise, bike sharing services will explode globally in 2018 as multi-modal transportation (for example, taking a train to the city, riding a bike-share to your meeting destinations, then taking a ride-sharing service to the airport) becomes more convenient and cost effective.

IoT trends: Security perspective from Thales eSecurity

These predictions come from John Grimm, senior director of IoT security strategy

1. Safety and security as one: As we look at how the IoT is, especially in manufacturing in which there are connected industrial-type systems, we’re starting to see how the operational world and the traditional IT world are coming together. Traditional safety such as protecting employees is now merging with IT security.
2. The changing face of vulnerabilities: Vulnerabilities in connected devices are becoming more wide-ranging, and with greater potential impact if exploited. The recently discovered “CAN bus” vulnerability, for example, is found in all modern cars, and could allow criminals to hack into a vehicle’s safety systems. Steps must be taken to secure these vulnerabilities the moment they’re discovered, before consumers lose trust in the IoT or worse…
3. Focusing on consumers: Product manufacturers—particularly of consumer devices—will continue to have poorly implemented security. Consumer awareness of security is growing but not enough to impact their buying decisions. This tide will turn, however, so manufacturers should be doing more to instill trust, or risk losing customers.
4. Getting IoT security right will lead to a significant competitive advantage: Security is not easy to get right. Encryption may be easy but protecting the key and managing it is not. However, companies that get security right and move out of the pilot phase will gain a huge competitive advantage.