The adoption of automation in business is now a question of when not if, but it’s not always an easy ride. Carmine Rimi, AI product manager at Canonical –  the company behind Ubuntu – outlines how businesses can mitigate some of the complications.

Businesses have always sought ways to become more efficient and drive cost savings. However, that search could be accelerated with the onset of Artificial Intelligence (AI) powered automation.

AI automation has the transformative power to trigger the fourth industrial revolution. The benefits of this technology are seeing organisations adopt it in droves; but developing and integrating AI models into a business can be challenging.

The rise of automation

Automation has had a massive impact on every industry from manufacturing to retail. For example PayPal, the secure money transfer service, successfully uses machine learning to fight money laundering via the analysis of millions of transactions between buyers and sellers.

It is examples of projects like this which are encouraging businesses to embrace new ways of doing things. Industries such as marketing have been at the forefront of embracing new technology, with automation well established among companies in the sector.

Sensape, a software provider based in Germany, is a great example of this trend. The company provides cognition to digital signage solutions by teaching the operating systems to see, understand and interact with delegates at trade events and customers in retail locations.

Through a mixture of AI, computer vision and augmented reality (AR), content is altered in real-time, enhancing the rate of interaction by 14.7 times when compared to traditional digital signage.

While the number of AI projects in the market is increasing exponentially, organisations in traditional sectors, such as finance, lag far behind in terms of achieving full automation.

Burdened by older technology and riddled with apprehension over the security of data, these sectors can experience huge challenges in embracing automation.

However, with competition increasing from cloud native financial companies, the onus is on large-scale financial organisations to adapt rapidly and deploy automation in order to compete more effectively.

While it’s good to appreciate the opportunities that AI can unlock, a successful deployment also relies on the ability to understand the challenges as well.

Jumping the hurdles

The in-house development of intelligent AI and machine learning models can often be very expensive.

The causes of these costs are normally having the compute power and data free to develop and program a sophisticated model.

The UK’s Digital Catapult Centre identified that the cost of a single training run for a machine learning system can exceed over £10,000.

Considering that this expense comes on top of the cost of storing a huge volume of clean and consistent data, it is not surprising that budgets can be one of the biggest blocks to embracing AI.

In spite of these costs, businesses are still clamouring to build their own AI applications. However, this does require fitting infrastructure to be in place to enable that development.

To build those AI solutions, minus the price tag, businesses are turning to cloud andopen source platforms.

When storing and computing in the cloud, organisations can leverage on-demand payment models and customisation to streamline investment in technology to fit to the capabilities that it needs.

Equally, open source models usually have limited or no cost at all and need less input from in-house developers. Open source, by its very nature allows businesses to build on publicly available AI applications to deliver the software that they require.

Developing skills

Although expensive, both data and compute power availability has increased over the last several years.

This has triggered a torrent of intelligent and automated services, which developers must pick their way through to identify the most suitable one for a successful implementation.

This goes beyond just identifying an AI platform which addresses business requirements. Consideration must be given to areas such as finding the right DevOps partner for consultation and support, particularly during the early stages of a project when staff members need to be trained.

Limitations on skills among the workforce must also be addressed if a business is to successfully expand its automation strategy. Investing in experts on AI is the best way to deal with this issue.

A recent report discovered that just 300,000 AI professionals exist worldwide. This means skilled developers have a monopoly on the market – enabling them to name the price of the services they offer.

Nevertheless, the global skills deficit is being combatted with cloud and open source solutions. Open source initiatives like Fast.ai provide a free open source framework which offers training in coding AI models, like image classification and natural language tasks.

It is a free tool that provides ongoing training, allowing in-house developers to retain the latest information, to the benefit of boosting innovation within the business.

Making the right decisions

While the IT team is at the epicentre of enabling adoption, successful digital transformation initiatives require a shift in attitudes across the business.

This is crucial to the success of automation, as the roles of employees are likely to be transformed. This makes their buy-in even more important.

At its best, automation is the linking of AI and robots with humans, to support on repetitive or strenuous tasks to boost productivity and keep employees satisfied.

Businesses which make the leap to automation record impressive results, with studiesrevealing that augmented organisations achieved 28% higher overall performance, an improved financial position (31%) and higher engagement among employees (38%).

Looking at these statistics, the question is not if businesses should adopt AI platforms, but when?

The application of automation to companies delivers a positive return on investment and improved productivity, enabling businesses to boost their ability to compete.

Similarly, an increasingly open technology landscape provides CIOs and CTOs much greater flexibility and agility when deploying cutting-edge technologies like AI.

They have greater access to the tools required to develop, adopt and innovate moving forward, which is essential to long-term business success.

The priority for business leaders must now be to adopt the mix of technologies needed to grasp that success, while still reducing the risks linked to these complicated deployments.