Digital transformation: Is your business ready for artificial intelligence?
![Chris Priest](https://www.ensono.com/wp-content/uploads/2023/03/Chris-Priest-150x150.jpg)
Chris Priest
Consulting Practice Director, Ensono Digital
![](https://www.ensono.com/wp-content/uploads/2023/05/image-11-768x512.png)
With a call for better, more immediate responses to customer needs, and the universal drive for operational efficiency and expedited data accuracy, Artificial Intelligence (AI) and Machine Learning (ML) is fast rising to the top of many IT leaders’ wish lists.
It isn’t all robot stock pickers and self-driving cars though (more of that later), as AI is already intrinsic to much of that which we take for granted. As more organisations are recognising the contribution that AI is making to their value, are you missing the opportunity to weave AI into your digital transformation plans? Or are you already on the path and looking to expand the role that AI and ML is taking in your business?
In this blog Consulting Practice Director, US, Chris Priest, looks at the opportunities and blockers for bringing AI into your operation, boosting innovation, and delivering benefits from cost saving to revenue growth.
What is AI?
Data is a vital part of any business’s digital transformation and AI provides a valuable way to unlock the benefits from this data. The immense volumes of information that organisations are now gathering, managing, and utilising need the right tools and methodologies to cut through and unleash the value. AI technologies provide some of the most powerful approaches to processing the intelligence that might otherwise remain trapped in your business and can use this data to learn more efficient ways of operating, provide tailored recommendations, and predict future states to keep you ahead of your competitors.
Artificial Intelligence refers to processes and algorithms that mimic human reasoning and problem solving, and machine learning – along with deep learning (DL) – are subsets of this concept. AI is not by any means a new concept, having emerged in the 1950s, but – with the benefit of flexible cloud-based solutions – has gained popularity over the last decade along with improved usability, accessibility and, of course, affordability. AI is more prevalent than many realise being evident in commercial applications such as internet search engines, streaming service recommendations, and voice recognition software. And AI investment is also reaping rewards in enterprise applications, using data to predict equipment failure in manufacturing, in smart home technologies, detecting patterns of fraud – or potential fraud – in banking, and even in the early detection of cancer.
Types of AI range from reactive machines – those that perform basic tasks in response to an input, with little to no ‘learning’ element – to ‘theory of mind’ applications that are able to infer something of the mental state of the humans interacting with it. There is a magnitude of variety, and a great diversity of complexity, within this paradigm. Machine learning (ML) lies at the heart of much of this difference. ML algorithms also cover a range of complexity from ‘supervised learning’, where the AI is fed specific data to learn from, and a model for the outputs, through ‘unsupervised learning’ to the most complex, ‘reinforcement learning’. Reinforcement learning works without a fixed data set but draws information from the environment in which it exists, receiving positive or negative ‘rewards’ for the actions it takes, over time AI driven by this form of ML will refine its responses to maximise these rewards. This type of ML can be seen in industrial robots, in operational management such as the Deepmind cooling processes in the Google Data Centres, and – at its most advanced – in self-driving cars. More of that later.
What are the benefits?
Following the COVID-19 pandemic AI benefits have tended, say McKinsey in their 2021 Global Survey on The State of AI, towards cost-saving rather than top line revenues. Indeed 27% of respondents to the survey reported that at least 5% of EBIT could be attributed to AI, up from 22% in the 2020 survey. In tangible terms, AI gives us the opportunity to deliver benefits across the three pillars of value, innovation, and service.
Bring AI into your business
From educating your stakeholders and operators, to calling on the experts for advice, introducing AI into your business needs to be a considered process. Of the steps that you need to follow to bring in AI capabilities, the two most important are to ensure that you are operationally prepared, and that your data is going to meet your needs.
BCG, in their piece on enabling AI returns, talk about the importance of operational adjustment. They report that businesses who have scaled AI across the organisation and delivered the greatest returns, invest twice as much in people and processes as they do in technologies. These are the businesses that have recognised and embraced AI as an integral part of their operations.
Data is the most important part of the foundation on which all your AI ambitions are built
At Ensono Digital we believe that the biggest enabler, and the most significant potential blocker to enabling AI in your business, is your data. Data feeds AI and trains machine learning. Poor management of your data can doom your project from the outset. Data needs to be clean, tagged correctly, and documented. Standards should be in place to ensure that your data sources will remain secure and consistent, and that collection, storage, and management is effective for your needs. Having the processes in place from the start that will annotate your data correctly – such as its meaning, or its sources – will help to line you up for success.
Blockers and opportunities
Volume, also, is key. The greater the volume of data the greater the opportunity to parse, process, and learn. And yet an exceptionally diverse volume does not necessarily result in a well-trained AI agent. Take, for example, a successful but fledging international used car sales business looking to develop a recommendations engine using AI and ML. Their huge volume of data, although highly valuable for BI and customer management, was too shallow an information set to fuel a valuable AI. Without the mass of customers over time and across brands, the ML algorithms were unable to generate meaningful relationships and determine, for example, that previous owners of a BMW 5 series are most likely to be upsold to a Porsche 911. Time (or a data purchase) is needed for the engine to deliver value.
Conversely, innovative tech/auto firm Tesla has generated one of the most impressive data sets to power their self-driving AI. Using deep learning overlayed on real-world data, collected from nearly 2 million Tesla Autopilot enabled vehicles, Tesla is in a position to lead autonomous driving technology – and its unique stance pivots on the type of data it has been collecting.
Although Tesla originally started – back in 2014 – collecting information to support its autopilot using Rradar, sonar and just one external camera, the design evolved through 2.0 with eight more cameras supporting the radar, and then ditching the radar altogether. Elon Musk’s logic behind the decision is all about the legitimacy of the data. When you drive, he says, you don’t use radar or sonar or lidar, you use your eyes, and Tesla’s cameras are even better than even the keenest human sight. Tesla has multiple cameras with overlapping fields of vision gathering data from a 360-degree view, on nearly two million cars, and uses this data to train neural networks to detect objects and measure depth in real time. They had over three billion miles of real driving data at the beginning of 2020, and some believe that surpassed 5.5 billion in 2021. It’s this data that puts Tesla in the race to be the self-driving leader – the technology just needs to improve its self-driving software to match the quality of its data.
BCG’s report states that with the fundamentals in place – technology, talent, and data – businesses have the opportunity to gain a 21% boost in ROI from AI initiatives. By layering on the operational readiness to iterate and build on those AI solutions ROI can be as much as 39%. Finally, with organisational learning, and the associated ability to evolve business strategies and processes alongside the AI, businesses can increase ROI by 73% and bring in significant returns from their AI implementations.
I say, do like Tesla, start with establishing your rich, detailed, reliable data then develop and iterate your solutions, engage your teams and roll out your business change, and watch the benefit of AI roll in.
About the author
Social Share
Don't miss the latest from Ensono
Keep up with Ensono
Innovation never stops, and we support you at every stage. From infrastructure-as-a-service advances to upcoming webinars, explore our news here.
Blog Post | September 5, 2024 | Technology trends
APIs and Abstraction: Modernizing with IBM’s z/OS Connect and Data Virtualization Manager
Blog Post | September 4, 2024 | Industry trends
Bridging the Skills Gap with SREaaS
Blog Post | August 29, 2024 | Industry trends