Bruce Zhou, CEO, Axilspot, explains how partners can overcome the challenges of IoT and the Internet worlds integrating.
The ability to connect machines, devices, sensors, and other everyday things into an intelligent network and make sense out of them, has huge promises to change everyday life. Manufacturing plants, retail outlets, households, residential and commercial buildings, logistic hubs, vehicles, roads, transit hubs, city governments, utilities, hospitals, healthcare, oil and gas fields, and scores of other areas are among the early adopters of these new opportunities.
As can be expected by taking in our entire world and attempting to change it one stroke and with one magic wand is unrealistic, challenging and cause for chaos in the short term. Some of this is reflected in many of the reports we receive on a daily basis on the Internet of Things (IoT).
For business decision-makers and technology providers, some measure of clarity and progress can be made by understanding the challenges of applying IoT to business. We look at two principal challenges in making this happen.
Technology protocols
The challenges of IoT are resident in the ability to connect sensors and devices directly to the Internet. To connect to the Internet, any device and application must use Internet Protocol stacks. It is for this reason that devices called Internet gateways have been created that connect to non-Internet Protocol devices and networks on one side and the Internet-driven applications, devices and networks, on the other side.
By doing so, Internet gateways need to be designed for the type of local networks and local applications that they manage on one side and how they connect to the Internet on the other side. By doing so they are less open to changes and adaptations of the sensors and devices for other purposes and applications, and limit their flexibility.
Over the years, countless Internet Protocol driven applications, software tools and diagnostic kits, have been developed and exist, that can be applied to IoT devices and sensors, with or without much adaptation. This will help to get the IoT environment up and running in a short period of time, relatively speaking. That is provided they are in the same Internet Protocol environment.
The basis for the development of Internet Protocol applications till now has been growing availability of large computing power and significant memory capacity to manage complexity of the Internet Protocol based applications. And all this within localised high cost nodes like desktops, notebooks, smartphones, tablets, among others.
This is the exact antithesis of the IoT environment as it exists today, which is low cost, low power consuming, low computing capacity, and distributed computing nodes.
Networking architecture
When building an IoT network it is essential to plan and look ahead. It is also important to converse with all departments associated with the benefits and operations of the IoT network and applications. When dealing with tens of thousands and hundreds of thousands of sensors and devices in any IoT project, it is important to have a networking strategy to ensure the end result is managed chaos instead of the opposite.
Industry consultants who have gone through such projects point out, IoT application and network designs need to be built looking at requirements five years from now. When in doubt the rule of thumb is to design networks for a requirement that is 10 times the number of devices that are connecting in now. Another important consideration is to design the wireless environment thinking about what will happen in that specific industry over next five years.
While the benefits of such best practices are self-evident, industry surveys show that such discussions are not happening as much as they should. According to Forrester’s Global Business Technographics Networks and Telecommunications Survey in 2015, only 45 percent of networking professionals have regular conversations with other departments. If all else fails, the minimum rule of thumb to apply, think big for IoT.
Looking ahead into the requirements for IoT type of networks there is a significant change in the way devices will communicate. More often than not, communication signals will hop from device to device, to enter and leave the network of devices, connecting into and out of their mesh network.
Using a number of hops between devices rather than the conventional hub to spoke jump will become the standard for most IoT applications and networks. The key here is multiple and random paths between IoT devices, based on predefined algorithms, rather than a fixed point to point hop. This saves power and cost while building scale and redundancy.
Such networks are called mesh networks and the devices are called wireless nodes. Since connectivity within mesh networks rely on each other, their efficiency grows with an increase in the number of devices. Mesh networks become more efficient as the number of wireless nodes inside them increases. This is the opposite of what happens with the hub and spoke, star network arrangement.
A mesh network cannot be easily shut down. To shut down a mesh network, every wireless node needs to be shut down. By comparison, if the hub is compromised in a star network the entire network can stop functioning.
The challenge and opportunity for partners with IoT lies in their ability to get sensors and devices connected to the Internet. Connectivity to the Internet requires applications built on Internet Protocol stacks. Moreover, partners should be able to design, build and implement, mesh type networks. This allows businesses to cost effectively grow large-scale IoT networks without incurring additional costs of bandwidth and connectivity.
As silicon technologies improve the computing capabilities of IoT sensors, devices, machines and appliances, they will increasingly become more and more Internet like. At that stage the two worlds will come together in exponential fashion and the true power of IoT will become a reality.
