What is User Interface Automation?

I originally found myself in a bit of a predicament this week, without a topic I really felt like writing about. I have to be honest, that does happen from time to time, but it’s always a bit of dismay to experience. However, the other day, someone asked me a question that actually was just the topic I needed. They had asked me, what exactly is user interface automation. After I, hopefully, had answered their question eloquently enough, they began asking me if it was beneficial to push for heavy use of the concept, and what the drawbacks were.

So, having drawn inspiration from that, I think it’s a good topic to discuss here, to be honest with you. User interface automation has always been something of some importance in software and web design, even back in the early days of visual interface. In fact, GUI is the direct result of one of the first pushes to automation by visually representing aspects of the system, as well as files and tasks in a manner that spares the user complex text commands. So, we all enjoy the benefit of a pretty big push to some level of automation.

The question was more about pushing further automation with today’s modern interface capabilities. That’s a really good question, and it wasn’t so easy to answer. Let’s look at what enhanced automation is by today’s standards before we jump to any conclusions. By today’s standards, a typical UI is not automated at all, as standard UI is just base level. Its ancestry in the command line era is no longer part of the measurement of progress in the topic. A few things have been famously tried, some of which have worked while others have not. Everyone remembers the legendary helper paper clip in the Office suite. This was an early attempt at a higher level of automation through primitive AI. It didn’t work well and just annoyed most, as did the search dog of Windows XP.

Later, we saw smaller, less brazen developments creep in that we honestly like now. Auto fill forms on operating systems and websites are much enjoyed most of the time. The same goes for auto correction in word processing and forms. Still, even this has an instance where it works in a calamitous way – we all know the infamously horrible things the iPhone’s messenger can make us say to our mothers, right?
By definition, the next push is to onboard systems that are somewhat self-aware and able to guide you with invitation in a deliberate way. These work well for tutorials and learning tools, and are often viewed as the future to practical self-service in SaaS. While these ideas work fantastically for some implementations, I have to say that perhaps some things are best kept simple. I don’t need Notepad to make suggestions on fonts and indenting when I randomly type in thoughts.

So, that’s user interface automation in a nutshell. GUI was the dawn of the concept, and then experimentation with it gave us little elements we take for granted now. The future, though promising, needs to be a world where everything has a place, but nothing like this is too ubiquitous.

For more information check out Graphical User Interface examples.