Kathy Gibson is at Fujitsu Forum in Munich – Robotics has evolved tremendously over the last few years – and cognition is now giving them additional capacity.
“We are seeing this over all types of automation,” says James Dening, vice-president and digital worker evangelist at Automation Anywhere.
The reason this is happening, he adds, is that the world is a hard place and businesses have a huge range of challenges.
Among the challenges are issues of more and different competition, demanding customers and increased regulation.
On top of this, IT complexity has increased exponentially – and much of it is unintegrated – while data continues to flood in from all directions.
“A lot of this data is hard to understand – just 20% of it is structured,” Dening explains. “Semi-structured and unstructured data is easy for us to understand, but it’s hard for computers.”
What we need to implement, he says, is intelligent automation. The transition from a human workforce to a mixed workforce is happening now.
“There are many things people are very good at. And there are things that computers are very good at.”
Indeed, chess grand master Garry Kasparov says the most formidable opponent he has ever faced is a reasonably good player backed up with a computer.
In companies, the problem with moving from robotic process automation (RPA) to artificial intelligence (AI) platforms is typically one of scale.
The problems that companies want to solve are probably too small or insignificant for the big AI platforms that are typically out in the market.
“If you are going to do cognitive automation, you need a set of tools as well,” Dening says. These need to overcome the challenges of recognizing, understanding, enriching and improving the information.
Dening says there are a number of best practices for successful cognitive automation. They are to identify high-ROI (return on investment) processes, select representative training data, then add additional processes.
“Imagine if knowledge workers no longer need to extract information from unstructured content – documents, images, emails and more.
“They can then be freed up to make decisions and handle exceptions.”
The change to cognitive automation will come quickly, Dening adds. “It is happening now. Every large company in the world has some automation project in the works. We will look back in five or 10 years’ time and wonder how we managed without automation.”