Digital transformation predictions for 2021

Gemserve offers some digital transformation predictions for the year ahead.

While 2020 was a year of innovation in digital services, the focus in 2021 is likely to be more about consolidation. After ensuring over the past 12 months that employees could work remotely, the next step is to keep users at the heart of digital strategies.

Migration of services to the cloud is nothing new and has been the bedrock of digital services for some time. However, with more organisations adopting cloud solutions and their users becoming increasingly distributed, a one- size-fits-all cloud solution can introduce latency and often falls foul of data residency policy. The traditional answer has been ‘hybrid cloud’ – a partial move to the public cloud but keeping high- value data within a private cloud.

Here, some of the advantages of the public cloud can be harnessed while maintaining some of the control provided by a private cloud. But, in the same way, that this provides the best of both worlds, it also limits the ability to take full advantage of the investment, skills and innovation that provide much of the value in the public cloud.

The new answer is ‘distributed cloud’, and increasingly vendors are providing more flexibility and scope as to how the location of cloud services is managed. For example, AWS Outposts extends the same infrastructure and services to any datacentre or on-premises hosting. Over time, it is likely that more specialised localised facilities will allow this cloud topology to be more distributed and – like wifi hotspots today – AWS, Azure, Google et al will have models that support distributed cloud.

Greater integration of technologies

Now that “the cloud” has entered common usage, the next phrase to make its way from the server room to the boardroom will be the mesh. The term describes the integration of people, devices, the internet of things (IoT), processes, services, and data into a collaborative environment – in other words, technologies increasingly working together rather than individually.

Though the mesh has its origins in the digital world it is increasingly being seen across other areas. For example, the energy mesh has involved small-scale power producers being integrated into the wider grid of large-scale power plants.

Feed-in tariffs for household solar energy producers is one step along this path. With the increasing availability of electric vehicles (EVs) and storage batteries, it will soon be possible for households to buy energy at low prices and sell it back to the grid at higher prices.

Devices need to be intelligent enough to know the current cost of electricity and to predict usage. A vehicle parked overnight or an empty house need to know when they can feed power back into the energy mesh.

The cybersecurity mesh involves a distributed architectural approach to scalable, flexible, and reliable control. Many assets now exist outside the traditional security perimeter, like smart house doorbells or cameras.

The cybersecurity mesh allows a security perimeter to be defined around the identity of a person or an object.

Instead of forming a perimeter around a location or ‘walled city’, policies will adapt to a ‘suit of armour’ around an individual.

Finally, the data mesh is a shift in how big analytical data is managed. Instead of having a single monolithic data lake, ownership of data is distributed, and so data governance models need to shift. Data analysis needs to be treated as a product that domains can share with each other.

Citizen programmer as the norm

While office-based functions have moved – with varying degrees of success – into a ‘work from anywhere’ culture during the past year, this has focused largely on ensuring workers have access to the right systems and the correct data to let them work from their own kitchen rather than their company’s office. Greater use of Zoom and Microsoft Teams has facilitated meetings, and greater use of online collaboration tools has improved interactions in the development of ideas.

If this move away from the office is to continue, then some of the ordinary tasks that are easier face-to-face need to be viewed through the remote lens. For example, the way teams are managed needs to change, for example, keeping track of who’s doing what, or getting ideas ready for the next team meeting.

These tasks can be accomplished via email, but it’s often time-consuming. In the ‘old days’, those jobs would have been completed quickly through a discussion in the open-plan office – shouting a question to a colleague is much harder on Zoom.

Writing programs to manage workflows, tasks, and activities used to take lots of effort, but tools are now available to help even non-programmers build such applications. This ‘democratisation of programming’ uses tools like Amazon Honeycode and the Power Automate function in Microsoft Teams.

Typically, low-code and no-code platforms operate in narrow and specific domains but come with a number of risks. Putting the power of these tools in the hands of non-programmers can lead to the development of applications without the design and rigour that software engineers bring – and if these applications are used in business-critical situations, then this lack of discipline can be a huge problem.

The responsible adoption of more sophisticated AI techniques

Artificial Intelligence (AI) and machine learning continue to show rich promise across a wide variety of sectors and applications. More and more, they are augmenting human decision making, and the insights can lead to direct action – especially where the decision making becomes fully automated.

This is most effective where the AI is closest to the work being done, operating in real-time and able to carry out decisions there and then.

Where this has a consumer focus – such as with Amazon’s Alexa and other personal assistants – this allows organisations to capture information, which they can then use to enhance products and offer better value to customers.

As the A-Level controversy in England showed over the summer, equations and algorithms need to be checked to make sure they aren’t biased. The governance surrounding the deployment and management of AI has been increasingly concerned with trust, transparency, ethics, and fairness, which has resulted in a more responsible use of AI, with greater visibility of how they are compliant with best practice to eliminate bias.

AI impinges on other digital innovations and will become increasingly used in technologies such as augmented reality (AR) and virtual reality (VR) to fill in the gaps when these innovations need to operate in real-time to support more realistic ways of working digitally. In engineering too, AI techniques are being used to provide more intelligent decision making when aligned to 3D-printing, adapting to variations in the physical environment and allowing for greater personalisation.

Using data to change behaviours

The IoT is allowing massive amounts of data to be collected across numerous sectors. This influx of data has affected everything from the way services are designed to the way business decisions are made.

The size of the IoT market is expected to grow from $20.2 billion in 2020 to $35.2 billion by 2025. That growth will be driven by a wide variety of factors, from preventing cyberattacks to boosting business productivity.

The next step is what has been termed the ‘internet of behaviours’ – using people’s behaviours to inform decisions. Once behaviours are understood, changes to them can be encouraged for greater benefit.

For example, the increased use of home deliveries during 2020 illustrates why it’s important for delivery companies to understand their drivers’ behaviour – how fast they drive, how aggressively they brake, which routes they take.  This information can be collected and analysed using big data techniques, with drivers given incentives to change their habits and save the company money.

Similarly, wearable devices are helping health and social care providers to monitor the health and behaviour of frail and vulnerable people, so they can live in their own homes for longer. A simple example is the use of smart energy consumption data to build a picture of what “normal” looks like for an individual, and then act on any deviations – if  Margaret doesn’t switch on her kettle at 6 pm like she does every evening then she may have had an accident and her home help can be alerted to give her a call to check she’s ok.

In a similar vein, wearables could be used to measure and monitor the behaviour of people working from home or in remote locations. If a company or an organisation can support its members of staff to work outside the office, then the explosion in home and remote working could be here to stay.