Autonomous Databases: Organisations are at risk of falling behind unless they make the move to automation, according to the majority of respondents to a recent Harvard Business Review Analytic Services survey and sponsored by Oracle.
It’s no surprise that some forward-looking organisations are already embedding artificial intelligence (AI) and machine learning (ML) technologies into their critical business systems and processes, with key areas of the business predicted to benefit the most from this type of automation being operations, customer service, decision support, IT and finance.
Yet, according to the report, many organisations are hesitant to make the move for a number of reasons. Not surprisingly, fear of the unknown is one of the main contributors and as with anything new, adoption takes time – companies need time to get their heads around how these emerging technologies can fit into their current enterprise systems and just how to do that within their existing budgets, skills and culture.
While this change certainly won’t happen overnight, respondents are expecting to significantly increase their use of intelligent automation over the next three years. With that the case, business and IT leaders need to start considering how to move along their automation journey from basic adoption to full intelligent automation systems. But how?
Six levels of automation
One way to do this might be to consider a model devised by Gartner around the ‘six levels of automation’. While specific to the supply chain, it is also considered highly applicable to the broader business.
The six levels can be seen as a path which organisations can consider, helping them decide where they want to be on that scale – both as an enterprise as a whole, but also around particular use cases or business processes. The phases gradually ramp up, starting with early, low-hanging fruit for intelligent automation around data-intensive and repetitive tasks that machines can do better and faster than humans, and moving up through stages where the system itself gains more and more autonomy before it can be classified as fully autonomous.
From looking at their business and its core activities as a whole, they can then determine what it means to them to operate as an intelligent enterprise, and how they can move from pockets of automation to a strategic, enterprise-wide approach.
So what are the six stages?
The six levels
Level one is the most basic form of automation – ‘give me the facts’, essentially asking the system to take general data around business areas like sales and production and analyse it before presenting the results.
The next stage, ‘give me a suggestion’, means the system can make recommendations based on specific information, this typically comes in the form of statistical forecasting.
After this the system begins to think for itself, where it starts to ‘help me as I go’. Here, automated alerts add a level of advisory guidance to the user as they plan and execute tasks.
The next level requires more of a reliance on the system with, ‘do this task for me’, now the system has to take the information at hand, make an assessment and then have the capacity, if told via an ‘opt-in’, to act on it. This could apply to tasks such as recommending the best approach.
At level five the system actually has responsibility for a task, such as automatic reordering ‘until it is told otherwise’. While at this fifth level this is still some form of human input, the final, sixth level is far more sophisticated – and the one that really pushes the boundaries.
Here the systems use a mix of machine learning and automation to display more human-like traits, so they can start thinking ‘autonomously’ for themselves.
An example of this would be the Oracle Autonomous Database, which is essentially a self-driving, self-securing and self-repairing database that automates key management processes, including patching, tuning and upgrading to deliver unprecedented availability, performance and security—at a significantly lower cost.
Autonomous databases have the capability to revolutionize data management, helping boost the speed of insight and drive significant increases in productivity whereby manpower can be optimized and resources can be deployed to higher value tasks.
While level six has obvious advantages with autonomous capabilities, even the early stages offer tremendous benefits to organisations. In addition to improving the efficiency and quality of processes, intelligent automation enhances decision making by automating the routine and learning from the large data sets enterprises increasingly have access to. This was named a primary goal of intelligent automation efforts by nearly two thirds of respondents.
In fact, with the massive explosion in data, it is increasingly clear that some decisions can’t be executed without automation, one respondent to the report says. For example, “any kind of real-time, next best offer, next best action kinds of things like what ads appear on which customers’ screens in digital marketing—that’s got to be done in milliseconds. Humans can’t think that fast and digest the information required.”
Intelligent automation addresses all of this, so it will be no surprise when the number of companies describing their enterprise’s use of AI and automation today as sophisticated and extensive will jump from just 10 percent to well over half within three years.
Before this can happen a substantial amount of change will need to take place, not least in the digital transformation of data, skills, processes, and culture. Fortunately, many companies have already embarked on their digital journey and by tracking against these pre-prescribed levels they can understand how best to take advantage of this emerging area and reap the benefits.
As a parting thought it’s always worth considering, that just because we can automate tasks, it does not always mean that we should.
Neil Sholay is the VP of Digital, EMEA, for Oracle.