5 challenges of intelligent automation at scale

In the mass production revolution of the 20th century, machines sat side by side with humans replacing toil and sweat. Today, machines (let’s call them bots), are now doing the same, only for our brains instead of brawn.

While clearly helping both businesses and employees break the shackles of mundane, time consuming and inefficient tasks, the boom in bots isn’t creating the wholesale changes some futurologists and doom-mongers would have you believe.

>See also: Robotic service orchestration: 10 ways it’s different to RPA

In fact, we’re talking about replacing tiny corners of human intelligence with very specific, task-focused bots. Some follow rules and replace fingers (RPA bots), others chat (as long as the conversation is pretty dry and focused), while some use experience captured in thousands of rows of training data to make decisions.

What can be automated, and when?

The upshot being that many companies are deploying new intelligent automation technology, beginning with robotic process automation (RPA). But this is far from a ‘cure all’ solution. Quite the opposite in fact because it brings a series of separate challenges relating to what to automate, and when, which are actually the key question businesses need to be addressing.

In some organisations, this can be a simple question to answer because there is one clear task that takes a lot of time and fits within the capabilities of the chosen RPA tool. However, in most others it is more complicated and requires much more consideration. When assessing what to automate and when, businesses need to answer five key questions:

1. What’s it worth? – This is about understanding all of the activities and processes that can potentially be automated in an organisation and, more pertinently, what is the value that can be extracted through that automation. At its simplest level, this is knowing each activity cost.

>See also: 5 steps to successful RPA implementation

2. Is it possible? – Namely, is the automation technology available right now to automate certain activities within the organisation? Of course, the answer to the question changes month on month, year on year as new technology is delivered into the market.

3. Can we do it? – Very often this comes down to complexity. There is clear evidence that automating large numbers of simple things delivers more value to an organisation than doing the same for a small number of complex processes. In short, if a business can deliver £1M of value by automating four complicated processes or 12 simple processes, most organisations will hit the £1M faster by choosing the 12 simple processes.

4. Is it sensible? – If a particular process is currently undergoing significant change – for example for regulatory reasons – then it isn’t sensible to start automating those activities when it is likely they will require reinvestment in two or three months’ time. Also, if there is variability within a specific process, for example, a task ranging from two to 20 minutes to completion, then it is likely there is a high degree of decision making in the process and without significant investigation, it is impossible to know if this variation is rules based or driven by experience. Either way, the whole decision process will need to be coded into or learned by your automation tool, which takes time and money.

5. Is it right? – It’s easy to deploy automating work that simply should not be done in the first place. Businesses know that automation resources are like hens’ teeth and getting the most out of them is essential. What is self-evident is that automating failure demand – i.e. automating work that is only being done because a part of the business failed elsewhere like logistics or customer service, is not smart use of automation investment. Neither is it a way to use these new and exciting technologies to deliver the best experience to customers.

>See also: AI and automation will be far more significant than Brexit

Finding the answers

It is perfectly possible to pay a team of consultants to analyse the processes and create a detailed assessment. The only problem is that it’s a one-shot piece of work and it’s expensive. Businesses can also try and create a repeatable way of collating this data from existing systems, however it tends to be either non-existent, inaccurate or difficult to access in these systems.

Smart businesses are turning to robotic service orchestration (RSO) as an answer. It is a new technology specifically designed to address the challenges of automation at scale, and answer the five questions above – although to explain how it does that is a different article altogether.

To quote a business evergreen phrase, “change is the only constant”. What might be the highest priority to automate today, might not be tomorrow as business priorities, demand and technology move the goalposts. Yet, in spite of this, these key challenges will remain for any business looking to automate at scale.

 

Sourced by Kit Cox, CEO, Enate

Avatar photo

Nick Ismail

Nick Ismail is a former editor for Information Age (from 2018 to 2022) before moving on to become Global Head of Brand Journalism at HCLTech. He has a particular interest in smart technologies, AI and...