The Robotic Process Automation Landscape at the Dawn of 2020
UiPath has left Automation Anywhere and Blue Prism in the dust while Microsoft has steamrolled into the fray having just announced one of its most exciting products ever!
RPA Rundown
Robotic Process Automation, or RPA, is not about replacing the humans behind the work. RPA gives businesses and organisations the opportunity to drastically save staff members time, raise their productivity, and increase their morale. This is because robots get staff away from the difficult, repetitive, and/or menial tasks they do every day and gets them onto more productive tasks. RPA can be used to augment humans in their everyday processes so that, together, the human and the robot can conduct tasks extremely fast with machine precision.
RPA is growing very fast. In this article, we will look at some of the biggest players in the RPA space. Blue Prism, Automation Anywhere, Power Automate UI Flow, and UiPath. Blue Prism is a UK company that has been around since 2001 and is one of the two main solutions that offers a comprehensive UI for building automation.1 Automation Anywhere was started in 2003 in California and takes a scripting approach to automation creation. UiPath was founded in 2005 in Romania and has since moved its HQ to New York to be closer to its main client base. UiPath is the other major player in automation that provides a comprehensive UI. Recently UiPath has been named the fastest-growing enterprise software company in history.
Microsoft Power Automate
It really is an exciting time for business and technology. In my view, RPA is the technology that will act as the Trojan Horse for AI adoption across business. Microsoft seems to have noticed this and cemented this idea by announcing its own automation software with AI capabilities called UI Flow, an extension of its Flow service. The whole Flow services have now also been renamed to Power Automate. This is significant because it not only gives its new platform access to over 250+ API’s for integration that currently exist with Flow, it also provides tight coupling with Flows AI builder and models, and Microsoft’s cloud cognitive services. This catapults the capabilities of Power Automate into the same realm of intelligent automation potential as UiPath.
For existing Microsoft customers the offering is that much more enticing. By providing Microsoft customers with a way to automate processes within their current Microsoft product stack a barrier to entry for RPA technology has been removed. That barrier was having to adopt a new technology suite and stack — Something inevitable with other RPA platforms.
Many features are being rolled out by Microsoft in early and mid-2020 and the actual automation is still basic compared to the other main RPA players — an example is that right-click is not possible on webpages. That said, if a process is simple and involves the requirements for API calls that Power Automate supports then I would still consider it as an option, especially since the pricing is comparatively very low.
UiPath and the Rest
Whatever you do on a computer and/or in front of a screen can be automated by RPA. By utilising AI technologies UiPath takes steps even beyond this. The software is pushing the boundaries of current AI and it’s helping to roll the wheel away from the theoretical world of data science and into the practical world of software development and DevOps.
UiPath is used by thousands of businesses globally, organisations such as the US Government are using it as well as many local Perth companies. The take-up by big companies and the promises that UiPath has made and delivered, particularly in AI, has fuelled massive growth. There are use cases for almost every industry with RPA technology.
According to a report by the Economist, 91% of organisations are using automation technologies. Of these, 51% make use of these technologies, and 73% of those claim they are ‘very’ or ‘entirely’ satisfied with the returns from automation technology to date.
In terms of AI, 69% of the organisations surveyed by the Economist consider AI capabilities a high priority and 79% think that automation works best when augmenting humans rather than replacing them. For UiPath, these statistics only bolster the sensical conclusion that the company is heading in the right direction. My favourite example of this is UiPaths recent addition of its new AI Fabric feature to its platform which streamlines the development and deployment of AI models alongside existing RPA processes.
Support
When you adopt a software platform into your business, you need to be sure that it’s well supported by the software vendor and the software community. This, in turn, means that support for users and developers will be there when they need it giving them a positive feeling about using the software and won’t leave them with a sour taste in their mouth every time they run into issues. UiPath Go allows developers to quickly find official and community built pre-made solutions to common problems which can help to speed up development time. UiPath also has an excellent forum community and formal training certifications via the UiPath Academy this ties in with the way that UiPath is committed to culture, which it attributes to a key component of its success so far.
Development
The method for actually automating processes begins with the processes themselves. They need to be well-defined and clear for the users and analysts looking at them before you can begin to look at the opportunities. From there, a good approach is to identify the subprocesses that are causing the most pain and those first. It can be tempting to look at the return on investment that automating each process can yield alone, but a holistic view of the processes needs to be taken so that other factors like morale increase, the time staff take to switch to the process when they go to conduct its tasks, and many others can be considered.
Once the targeted processes are defined, the RPA consultant would then go on to work closely with the people who conduct them to learn their ins and outs. The consultant will try to understand how the process is carried out physically by the user, but also what the user is doing cognitively. Most RPA platforms show their weakness when it comes to replicating the cognitive functions of humans, but that’s where AI comes into play. Luckily, UiPath allows the RPA developer to integrate AI models and services into their RPA solution which gives them the tools they need to replate many of the cognitive functions that humans would usually need to step in for. As a backup, UiPath also makes it possible to hand over to humans if they need to, for example, to solve a very difficult captcha.
Architecture
All three RPA software solutions are able to use a client-server architecture where a server manages client bots on remote machines. UiPath can be deployed web-based solution where the Orchestrator software holds all login and log data and manages the machines and robots across the RPA ecosystem. This can be deployed on-premises or in the cloud and is highly extensible. Orchestrator is able to store processes agnostically and dish them out to available robots on available machines whenever there are tasks that need to be done. Naturally, this can be achieved in a queue like fashion or in a more manual pattern where orchestrator can be told exactly what to do via API calls.
Build Experience
When you sit down to actually build your automation process the experience can differ between the available technologies. You cannot record sequences with Blue Prism, however with Automation Anywhere you can record basic macro actions. Automation Anywhere is geared mainly towards developers and features a script based method of defining your automation sequence, therefor, coding is mandatory. UiPath and Blue Prism both have drag and drop UI that allows the user to easily define their process in a visual way. UiPath offers the added benefit of allowing the user to record their actions into a process. This can allow BAs with an understanding of the business process to create very fast prototypes which can be fine-tuned later and configured for scalability and best practice by developers.
AI Integration
RPA is a great segway for businesses of all sizes to implement AI technologies. The exciting part is that once more businesses are automating their basic processes, they are able to expand into more complicated processes using AI technology.
UiPath is taking the sequence definition of RPA to the next level using AI technology. Just by fulfilling your process multiple times in front of the watchful eye of UiPath’s inbuilt AI, you will be able to train it to know how to layout the sequence of your process. It will do this for you, and then leave it to you to conduct whatever touch-ups or additions you like. This is a visual process conducted by the AI and will lower the need for technical selectors and APIs in the RPA sequences which can be prone to errors. It will also leave UiPath robots with an enhanced ability to handle unexpected occurrences such as errors, modals, and front end/backend software upgrades.
This is not the only place that AI is being applied in UiPath. There is already the use of powerful image recognition and OCR available within the platform. Deep Learning techniques, including Image Recognition, can allow robots to conduct tasks that were previously exclusive to humans such as reading documents, viewing images, and even processing their meaning. Robots can communicate using Natural Language Understanding AI technologies, reading human text and even responding to humans via email and chat. AI Machine Learning techniques such as Anomaly Detection, Clustering, Categorisation, and Regression algorithms can all be utilised in order to analyse and optimise processes and provide insights like never before. Not only can this be achieved as API calls from the RPA solution, but using UiPath’s recently released AI Fabric, developers will be able to integrate their own models straight into UiPath sequences.
Traditionally you would need a lot of data for AI to begin to show a return on investment, planning and execution would usually require a Data Science team. Because the barrier to AI implementation is high, the original costs of implementation can be high. When businesses have already allowed UiPath in to automate processes and are seeing the benefits, AI is naturally introduced and can be built upon from assisting the RPA solution, to providing further analytics. This is the way that UiPath acts as the trojan horse for other pragmatic AI technologies.
Licensing
In terms of licensing costs, Automation Anywhere and UiPath both have community editions. This is a major benefit for companies that are looking to test out new technology and conduct R&D. UiPath’s Community edition is free forever and provides an easy upgrade to its enterprise edition. Blue Prism requires a licence even to try it out, or you need to be enrolled in the academy.
As you and the people around you work each day, think about whether you could be working smarter by using RPA technology.
An RPA partner, such as XenoVis AI, can assist you in discovering the viable RPA opportunities for your business and enabling you to go the full distance to implementation.
