📣   AKIPS Announces Acquisition by Tufin.   Read More

Blog

IoT Network Management

IoT network management: billions of devices and billions of dollars

Cast your mind back to 2000. Almost every device connected to a computer network was hardwired. The first WiFi devices were introduced in 1999 when Apple released its Airport Wi-Fi router (with one of Steve Jobs’ famous ‘One More Thing’ stunts at Macworld) and started incorporating Wi-Fi into Macs. Internet access first came to mobile phones at scale in 1999 with Japan’s NTT Docomo launch of its i-Mode service.(Nokia launched its Internet-enabled 9000 Communicator in 1996, but it was very much a niche product). Corporate networks were primarily used to connect desktop computers and servers with, in some cases, internet of things devices such as sensors, and some organisations offered dial-in access for employees.

Today, almost every computer, mobile phone and every Internet of Things (IoT) device associated with an organisation has, potentially, continuous wired or wireless connection to its network. And IoT devices are driving connections to numbers never previously envisaged.

The world’s population today is about eight billion. Let’s be generous and say there are two connected devices for human use (eg mobile phone, tablet or computer) for every person on the planet. The number of IoT devices is already at that level and is expected to double to 29.4b by 2030, according to Statista. However, estimates of current and future numbers vary enormously. In 2020 Juniper Research put the total number of IoT connections at 35 billion and predicted 83 billion by 2024.

Needless to say, this enormous number of connected devices creates potentially enormous network management challenges and a potentially enormous market for solutions to these challenges. Future Markets Insights in late 2022 estimated the market for IoT network management systems to be worth $US5.1b ($A7.5b) and to grow spectacularly: at an average CAGR of 23.3 percent over the next decade to reach $US41.5b ($A61b) by 2032. It identified several distinct network management functions driving this demand: configuration, topology, security and fault management and IoT network maintenance.

Network management aside, the scale and complexity of IoT networks have given rise to a related but distinct management challenge, and hence a product category: IoT device management.

IoT network management focuses on managing the underlying network infrastructure, whereas IoT device management is all about managing the individual devices that connect to the network. They are both essential to ensuring an IoT system is secure and operates optimally.

It’s difficult to know how much overlap there is in market growth forecasts, but those for IoT device management are even more spectacular than for IoT network management. According to Grand View Research: “The global IoT device management market size is expected to reach $US19.77b ($A29.00b) by 2030, registering a CAGR of 34.9 percent from 2023 to 2030.” It valued the market in 2022 at $US1.88b ($A2.76b).

TrustRadius lists 47 IoT device management products, which it describes as products that “allow businesses to remotely manage their fleets of IoT devices …. [providing] provisioning and authentication of IoT devices, remote configuration and management, data collection and reporting, real-time monitoring, and over-the-air software deployments for updates, patches, device onboarding/offboarding, etc.”

However, the primary challenge of IoT network and device management is not one of scale, it is one of complexity, and of simplicity. Many IoT devices are passive sensors that send small volumes of data from time to time, so real-time connectivity is not critical, nor are sudden surges in traffic volumes an issue for many IoT devices. But their simplicity can be their Achilles Heel. It can make it difficult to secure.

In one notorious incident, a casino in the US had a smart thermometer in its aquarium connected to its main computer network. Hackers could compromise the thermometer, gain access to the casino’s main network and exfiltrate its database of high rollers.

This incident exemplifies a widespread issue with the Industrial Internet of Things (IIoT): the billions of devices used for monitoring and control in industrial environments. In the past, such operational technology (OT) networks were rarely connected to general purpose corporate networks or to the Internet, so security was not a major issue. That is no longer the case, but many OT devices use proprietary and often use legacy software, making them difficult to secure and to monitor for compromise.

Securing IIoT networks could be an application well suited to artificial intelligence. Without having detailed knowledge of the myriad IIoT devices on a network, AI can build up a knowledge of normal behaviour and recognise anything out of the ordinary that could be a sign of compromise.

In fact, according to ARC Advisory Group, “the confluence of AI and Industrial IoT technological forces gives rise to a new digital solution category – the Artificial Intelligence of Things (AIoT) … [that] describes the combination of AI technologies with the Industrial IoT to enable the next generation of Industrial AI infrastructure, allowing organisations to enable seamless human-machine workflows, harmonise industrial data management, and rapidly transform raw data into tangible business outcomes.”

“Seamless workflows,” “harmonised data management” and “tangible business outcomes from raw data” are inspiring and visionary concepts, but they will not be realised if, at a much more basic level, huge IoT networks with hundreds, or maybe even thousands, of different devices are not kept secure and operational. A prerequisite for the realisation of those visionary concepts is likely to be the use of AI to manage such massive networks and their many different devices: some very intelligent and others very dumb.

Network engineers love AKIPS. Find out why

Get started with a free 30-day trial. You’ll receive a free demo too, before you deploy our software.