One of the things that we discuss often in the area of robotics process automation, or RPA, is that it’s a platform, not a product. You can’t just buy a full RPA solution off the shelf and automatically achieve enterprise automation.
There are a couple of steps that we follow and recommend in how to approach the objective of establishing a next-generation automation platform. The first thing that is important to understand is that everything must have an identity. We don’t really manage identities in IT the way we think we do. We think of identities separately: there are people in Active Directory, machines have host names and IP addresses and they’re managed some other way, and those devices are part of applications and software and we track that a different way – and all these things are associated with locations and we track those in different systems. The first major paradigm shift means understanding that identity is the base of all automation. Everything needs an identity and they need common, interrelated identities. The first step is to establish a global identity management platform that understands people, machines, locations and software.
I want to emphasize that RPA must be thought of as a platform and not a tool. You must create the mechanism to mesh together people actors, machine actors, client systems and partner systems. This is accomplished through a digital services integration platform that supports application program interfaces and web services, and facilitates the macro-level delivery processes.
These are two of the foundational pieces that need to be put in place as the foundation of your platform. Third, you need to ask, “what are the automation tools we use today?” You have end-point devices, servers – there are devices all over the place. It is important to understand what you have and establish a common way to drive micro-orchestration at the device level. We gave everything a name, we created a meeting place for everything – the digital services integration platform – and now we implement tools that know how to manipulate the objects that we’re trying to automate.
Today in IT, there are no truly universal infrastructure strategies. Most organizations have acquired different tools for different platforms. Each is managed in its own way, often vertically, with tools from various partners. IT has grown up with these specialized automation tools and the challenge is that we consume IT horizontally, not vertically. You need to be able to work horizontally on the network, the device, the server, across them all. A clear, integrated strategy for automation is needed.
Next, we need to attend to the ugly matter of data. Do you have accurate data about your machines? Consistent data about your software? Do you have an understanding of which pieces of software are installed on which machines to deliver system resources? How do those resources join together to deliver applications that the businesses consume and rely on? Getting that data and moving it into the right forms of representation are a critical path to automation. We’ve been living in the world of structured data – rows and columns – for many years; however, those forms of data representation won’t effectively support machine learning, leaving us with a need to establish a new generation data platform.
Once the building blocks of your platform – an identity, a way of interacting between machines and humans, and ways of manipulating the machines, and data – are in place, you are positioned to implement an RPA-capable automation platform.
CXOs aren’t going to trust their IT shops with a “Big Bang” approach to master automation. There’s a belief that humans are still better able to cognitively think through problem states. Recognizing this, a rational next step is to integrate our RPA-capable automation platform into your existing event management technologies, in a way that enables humans to delegate fault events to an RPA platform, on demand, in order to observe successes. This way, the operations people: 1) begin to trust the automation; and 2) they are there if the automation fails – and they know what additional knowledge needs to be fed into the automation platform so it doesn’t fail in the future.
RPA reduces the number of people who are sandwiched in the middle, administering and operating data center machines. Part of the ultimate answer is to extend the interface for the RPA-capable automation platform to each end user – we must force ourselves to think about new ways and forms of digital engagement to accomplish this. We have to rethink how we engage with the end users and how we get from “IT speak” to very simple, actionable, informative and even predictive digital representations for users, through portals and web apps.
The final step in this exciting journey is to take that engagement all the way down to the end user’s device level, so it’s on your smartphone, your laptop, your tablet. We do this by creating device-level apps that are equally as smart and understand how to take all the automation capabilities in an expansive IT infrastructure and deliver a simple, “touch and swipe” interface that enable us to be more productive.
Automation is only a small piece of the overall, expansive IT infrastructure. You don’t buy a tool and find yourself ready to go. We’re talking about a set of technologies and a methodology for implementation so organizations derive value out of every step of the journey. Workshops, surveys, data collection and implementation design are all necessary to navigate this journey.
In future posts, I’ll discuss each of the key steps and why they’re important. In the meantime, I look forward to your comments and questions.