Future of IoT: 6 trends and predictions to watch

Although challenges with IoT security and data privacy persist, IoT will continue to grow, propelled by targeted business goals and AI for data analytics.

Just like people find it difficult to picture a world without cell phones, a future where objects aren't connected to the internet is increasingly difficult to imagine. Manufacturers have already incorporated IoT in smart home devices, cars, cities and factories, and experts predict their use will continue to grow.

IoT has made it through its emerging years where organizations wanted to be part of the hype but didn't necessarily know what to do with the technology yet. Now, organizations begin IoT initiatives with specific objectives to fulfill and even more effective supporting technology.

Sensors cost less and organizations have many options to connect, whether that's Wi-Fi, low-power WAN or cellular. Greater embedded processing power fits into even smaller devices, and data processing can take place throughout the data pipeline from the sensors to the cloud. Orchestration management frameworks and platforms have given organizations the ability to analyze large quantities of data.

"It's going to be just like when we used to go to conferences, and it was a novelty to see a connected fill-in-the-blank. Look, it's a connected water bottle; it's a connected toilet. Now, if you go to CES and something's not connected, it's an anomaly," said Christian Renaud, research director of 451 Research's IoT practice. "When we have that [connection] now, we can't ever undo that."

IoT will continue to grow at a rate higher than 30% over the next few years, said Alfonso Velosa, research vice president and analyst for IoT at Gartner. Researchers predicted that by 2025, the total number of devices is expected to grow to 75 billion, according to Statista Research Department, or 80 billion IoT devices, according to IDC.

IoT will continue to grow in device numbers and use cases, but organizations must reckon with the security and interoperability challenges that have plagued the market since the beginning.

Emerging technologies and trends affect radar -- internet of things

1. IoT projects will be driven by business goals, not technology

Organizations have passed the days of buying IoT because of the hype without a specific goal. The future of IoT requires organizations to focus on the team or process in need of something, whether that's an IoT application or product to track equipment. For example, a hospital might add IoT sensors to locate wheelchairs or measure a patient's gait while they walk.

There really isn't an IoT market, but rather vertical markets that use IoT for business outcomes, Velosa said.

IoT will be part of people's jobs but not necessarily a job focus itself. Instead, IoT is the way organizations complete objectives. For example, an organization might want to drive workforce productivity and optimize the supply chain. IoT devices can help collect the data to find the ways to improve efficiency. The business teams and IT personnel would interact on a project that uses IoT.

2. IoT architectural groundwork will create more use cases

For IoT to thrive, vendors must step forward to connect the technology that makes IoT possible. Vendors might sell a piece of capital equipment as a service, but that's only possible if the vendor has all the connectivity, sensor data, storage, analytics and hardware in place.

Through 2025, large organizations will have, at least, a dozen IoT platform vendors, which means that they'll need one ring to rule them all.
Alfonso VelosaResearch vice president, Gartner

IoT deployments have a few horizontally applicable use cases, such as asset tracking or video analytics. These core use cases share a common IoT framework that a vendor can take and apply first throughout an industry vertical, then apply it to adjacent industry verticals. Likewise, an organization could start their IoT initiative with asset tracking with a fast ROI and, once the organization establishes the IoT architecture, they can get creative and apply other use cases. For example, automotive organizations can incorporate the connectivity and sensors in their vehicles and turn those features on or off for customers with a switch in the software, Renaud said.

"That's going to be a lot of fun, because we're starting to see a lot of people getting [into IoT] with asset tracking and video analytics today. And putting all this infrastructure in that they're able to leverage for every single additional incremental use case," Renaud said.

3. Interoperability issues will eventually winnow vendors from the market

Organizations will continue to face interoperability challenges with vendors, either using available integration capabilities and device management or building their own and supplementing it with open source capabilities.

The IoT market has had a multitude of vendors that use different architectures and protocols, but the market has started to cull the vendors, protocols and standards most widely used. Data acquisition and protocol translation and standardization platforms can translate legacy protocols to communicate with proprietary IoT services. Many manufacturers develop new devices with IP, which has the capability to speak in the evolving protocol stack.

But organizations will always have legacy equipment that uses different protocols. In the future, they will likely have an IoT gateway plugged into them to mediate communication. IoT platforms that analyze data grows more concentrated with hyperscale cloud providers, such as Microsoft, Amazon, Google and IBM. The future of IoT is moving toward all devices interacting and interoperating at a data level for analytics. Customers won't buy the products that don't interoperate or speak the language of their other devices and services, which will force some standardization of interoperability.

"Through 2025, large organizations will have, at least, a dozen IoT platform vendors, which means that they'll need one ring to rule them all. At some point, they'll have one cloud or other partner to plug into all their different IoT platforms," Velosa said.

4. IoT data analysis will inevitably lead to AI

Some of the biggest spending in infrastructure over the next five years will be on AI and machine learning, which are just degrees of analysis sophistication, Renaud said. Organizations must manage and analyze huge amounts of IoT data. They will start with machine learning now and it will grow and erode in favor of AI. Traditional analysis will fade away and organizations will increasingly use AI.

"It's sort of like the internet-connected toilet at CES. Everything's going to have connectivity. Everything's going to have sensors. Everything has to have some form of AI," Renaud said. "That's how the people who sold it to you are monetizing it -- as a service -- or they're doing secondary monetization of the data of your usage of it."

5. IoT will continue to be fraught with security challenges

Security is considered the No. 2 barrier to scaling IoT deployments, but vendors become incentivized to plan for security if the effort will get them the business deal, Velosa said.

Customers want to add security, but when they find out the cost for it, it gets pushed off. Although vendors invest in security, they don't always plan out the whole architectural view or threat surface topology, and not all vendors or partners have the right security skills.

Also, organizations often scale without fully putting in the security measures necessary. The future of IoT will always have a security element in it, so IT administrators, CIOs and the business team must work together to ensure they drive and support the policies and procedures for IoT security and push vendors to develop better security practices.

6. Privacy concerns continue until precipitating events push greater regulations

Many vendors have developed business models around monetizing data. Unless more regulations on monetizing data develop, this trend will continue. Organizations have more awareness and sensitivity around their data than an average individual customer, and more regulations have already been put into place for businesses, such as data sovereignty laws.

For example, if a vendor sold a robot as a service to a factory and must pull data off it for maintenance, then, by that country's laws, the telemetry data must stay in the country. Even with a limited amount of data from a robot, people can infer information that organizations might want to keep private, such as calculating the number of vehicles manufactured in a year based on the number of times the robot performed a specific function.

When organizations don't want information known, they must work with vendors to limit the collected data to the bare minimum of what the vendor needs to operate its service. Edge computing can also keep more data in an organization's network. Major regulations to protect data will require significant precipitating events, such as a cyberattack that compromises customer and proprietary data.

Dig Deeper on Enterprise Internet of Things (IoT)

SearchCIO
SearchSecurity
SearchNetworking
SearchDataCenter
SearchDataManagement
Close