By 2025, the worldwide web of today will look quaint. A new group of users will have arrived that vastly outnumbers human activity and bandwidth consumption.
And that group is the machines: devices connected to the internet and uploading information from all over the world, from across all sectors, servicing both consumers and businesses and governments alike. These devices, in many cases, will also download information, either to update their own software or to carry out a physical act from a centralised command.
This is the Internet of Things (IoT). And it will transform every type of industry.
In cities and towns, Flow Rate meters will monitor the water rates, helping to provide live data of drainage systems and rivers. In traffic, using a combination of cameras and data from traffic control systems, and increasingly from the ever smarter cars themselves, bottlenecks will be identified at certain times of day and damage reported in real time. Pollution levels and local weather would be tracked by air quality and atmospheric measuring devices, enabling extremely local weather forecasting in real time. Energy distribution and generation, particularly if batteries become an increasing part of the energy mix, will be controlled by data generated from the IoT, matching supply to demand. Drones will monitor infrastructure, looking for signs of damage and fallen cables on power lines.
In healthcare, portable and wearable medical devices, such as a digital glucose monitor for diabetes, or pacemakers for heart conditions, will also record data and send it back to the medical provider in real time. In manufacturing, devices will simply make more devices. In agriculture, robot fruit pickers will patrol the fields and, in some cases, perhaps manage the growing cycle entirely.
In our homes too, we might let a delivery driver enter our property through a camera-enabled smart lock, vetting them on our mobile from miles away. We can turn our heating on and off before we leave the office, and before we even leave our bed or reach for a remote control we can ask a digital assistant about the weather outside, or to read us the news or play our favourite song or remind me of my to-do list.
Many of these technologies are in use now, and are available to most people, and there is no reason to presume that these technologies won’t become more widespread and more integrated.
And the one thing they all have in common is the vast data streams they will generate. Many of these devices will operate all the time, recording and broadcasting information. Many of them will also receive information which will relay a new instruction, perhaps for an aerial drone inspecting infrastructure to examine a certain area in more depth. Furthermore, much of this data will be in video and audio format, which will probably be the devices that consume the most of any data package.
This data generation presents a problem for engineers and those who provide IoT services that needs to be overcome: how to manage the bandwidth to achieve the connectivity necessary. Bandwidth is a resource in increasing demand at a time when the costs of downtime, or connectivity failures, are increasing. This twin pressure to get connectivity right and to secure uptime is at the forefront of any IoT rollout.
And already, the IoT is causing headaches amongst CIOs and CTOs. With data scientists and programmers taking the safe approach and feeding back as much data as they can to the organisation’s servers, a debate is needed about what data should be transmitted and why. Bandwidth use might have to be reserved for important and actionable data only, whilst secondary data might have to wait for units to be manually checked (such as a local hard drive changed over), or downloaded if the device is returned to the organisation (such as an aerial drone’s flight data, rather than the more important information about the infrastructure it was used to check).
The arrival of 5G will certainly help the IoT in its rollout, but it isn’t an ultimate solution, and 5G itself will have to be underpinned by a fibre skeleton. Edge Computing too, where more data is processed ‘at the edge’ of the Cloud rather than being uploaded in its entirety, will also help to alleviate the bandwidth stress. But there is one more key technology that we believe will help manage IoT bandwidth consumption still.
That is bandwidth on demand. Controlled by Software Defined Networking, and with the ability to upscale or downscale your bandwidth configuration and change its limits in near real time, it is a key technology that we think will underpin the IoT revolution. Without this ability of flexible bandwidth on demand, then the IoT may never realise its full potential.
And bandwidth on demand is not just useful for providing more or less bandwidth, it also allows for clever use of that bandwidth. Via the SDN technology that underpins it, it is possible to prioritise certain traffic over others. Identifying data priority and setting rules for origin devices will lead to the development of a more streamlined IoT that makes economical use of bandwidth and reduces the chance of any interruption to connectivity. On top of the ability to schedule routine data transfers to and from the device network, such as a software update, bandwidth on demand is certain to play a key roll in the future of the IoT.