IoT, AI and big data require more than just cloud services

Nowadays, many organizations choose to house their ICT in the cloud. This makes sense as many applications are now available 'as a Service' and there is also a wide range of cloud services available for infrastructure. This allows organizations to benefit from market expertise, faster time to market, easy scaling up and down and less upfront investment. Organizations can focus more on the functionality (the what) and less on the technology (the how) of their ICT.

The rise of IoT, big data and AI enables all kinds of new and innovative solutions. An example is a building in which climate control and lighting are determined by algorithms based on data collected by a large number of sensors. This will also create IT landscapes within organizations in which smart sensors, devices, data centers and the cloud are connected and intensively exchange data to perform automated tasks.

With some of these types of automated systems, for example, because they are real-time, data transmission must be guaranteed and there should be little delay. A self-steering vehicle must decide in a split second whether to brake or accelerate based on information gathered by its sensors. A mobile heart monitor should immediately sound an external alarm or take action the moment its sensors register no or irregular heartbeat.

In cases like this, central data processing based on the cloud services now common and also a central data center often falls short. Due to a multitude of factors, data transfer may not succeed immediately or the required response time may not be realized. Consider limitations in mobile coverage, congestion on the Internet, the distance that must be bridged within the network or a ddos attack on the cloud provider's servers.

For this, edge computing offers a solution: computing power present close to the sensors and devices does the processing of this information and ensures the automatic execution of tasks or actions. Here, edge computing can be realized in a variety of ways. Think of the computer on board the self-driving vehicle, or computers in an equipment room at a hospital location, but also computing power at the edges of (mobile) telecom networks. There is no one-size-fits-all here; the situation determines which solution fits.

It is expected that due to developments in the fields of IoT, AI and big data edge computing, the majority of data, i.e. more than 50%, will be processed outside the central data center or cloud in the coming years. It is not for nothing that cloud providers (including hyper scalers) and also large infrastructure (software) vendors are expanding their product and service portfolio with various edge computing solutions. These solutions are often still vendor-specific and therefore integrate well with the cloud services or infrastructure products offered by the party in question. It is therefore particularly important to make a good consideration when choosing a product or service.

If your organization wants to be innovative in the future, it is wise to start thinking about the possibilities and necessity of edge computing now. Does or will your organization use IoT, big data and AI? Does your organization follow a "cloud unless" policy or are you moving to bring everything to the cloud? Do you automate processes where response time counts? If the answer to any of these questions is "yes," then it's time to rethink your IT strategy and incorporate and work out the opportunities, constraints and implications of edge computing.

Want to know more about edge computing in combination with IoT, big data and AI? Highberg (formerly known as VKA) advises various organizations on these topics. We are happy to come and talk to you.