The changing face of IoT connectivity is driving increased requirements for device management. This article examines three of them: increasing security needs, the rise of LPWA technologies and growing complexity of application implementation.
The first trend that complicates device management is the increase in security requirements for IoT devices. In a way IoT is a victim of its own success. Just ten years ago there was barely a billion devices that could be described as IoT. Today that figure is around 9 billion, and by the end of 2030 it will have risen to 25 billion. With growing numbers comes an exponential increase in interest from hackers.
There is also a growing raft of regulations including in Australia, the EU, Japan and the US, relating to IoT security. While most of them are positioned as voluntary, the likelihood is that meeting regulated security requirements will become increasingly important, either because of risk of sanctions against the manufacturer or by virtue of increasing user demand for compliant solutions.
Security demands also needs to be balanced with the need for products to be user-friendly, profitable and quick to market. Add in the need to maximise battery life in some cases (see below) and you have a very complex set of interrelated demands. Layering on security protocols which will be good for the lifetime of the device will be increasingly challenging.
The second trend is the growth of Low Power Wide Area devices. Transforma Insights estimates that by 2030 there will be 4 billion LPWA connections (up from just 220 million at the end of 2019). The sheer volume of connected devices clearly creates an elevated scale of requirement for device management, but more important are the characteristics of these technologies and the applications they will connect.
LPWA devices can support only limited downloads, necessitating a rethink of over-the-air device management. There is a lot of variation between the technologies, but all of them share the same principle: to extend battery life by reducing the amount of traffic being sent and received. For Sigfox the effective data transfer capacity is around 1KB per day, making firmware updates nigh on impossible. But even for higher functioning technologies such as NB-IoT and LTE-M there is still a big incentive to reduce the amount of traffic being sent and received so as to maintain long battery life.
Furthermore, LPWA technologies are overwhelming used for applications where human intervention is unlikely. Most IoT devices today are accessible by a human that can, if necessary, stage a manual intervention to correct the device. Major legacy applications, from fleet management to industrial SCADA systems, are higher value and usually relatively easily accessible. This makes them less painful to have to manually manage than billions of remote environmental sensors which individually don’t justify a truck roll to reset and are almost always in highly distributed and hard-to-reach locations.
The final trend is slightly more nebulous and relates to the changing nature of application implementation. One aspect of this is the requirement for more efficient provisioning. There is an increasing requirement for applications to be deployed into the field and work without the need for manual intervention. This low- (or zero-) touch provisioning is essential for many IoT applications to be cost-effective. It’s a large part of the reason why there will be significant demand for eSIM – not having to mess around with swapping out physical SIM cards. However, localisation is not simply about SIM cards. It also relates to numerous other functions of the application and the device. Different countries have different regulatory environments relating to data sovereignty, for instance, necessitating different approaches. Also, data delivery (e.g. APNs) needs to be configured. All of these things can be done manually, but IoT applications will increasingly rely on zero-touch provisioning, meaning that efficient OTA device management is a must.
The other big challenge related to application implementation is edge computing. There has been a noticeably trend in the last few years of more movement of application processing, including quite sophisticated elements such as machine vision, to the edge device. IoT devices are becoming much smarter. To quote Bob Swan, CEO of Intel: “increasingly everything looks like a computer”. There is a lot of truth in that, but unfortunately unlike computers, most IoT devices don’t have a convenient human to reboot them when they go wrong, or implement patches. Putting more smarts on unmanned edge devices creates a greater device management headache.
This article has covered just a few things that will drive additional device management requirements. Learn more about the changing nature of device management in IoT, and specifically smart cities, by joining the ‘Cellular IoT for Smart City: An Evolving Landscape’ webinar on 19th January at 5pm CET. Matt will join representatives of 1NCE, IoTerop and Itron to discuss the use of cellular technologies for smart cities and the impact of device management and LwM2M in IoT.