Join Transform 2021 this July 12-16. Register for the AI event of the year.
Robotic process automation (RPA) — technology that automates monotonous, repetitive chores traditionally performed by human workers — is big business. Forrester estimates that RPA and other AI subfields created jobs for 40% of companies in 2019 and that a tenth of startups now employ more digital workers than human ones. According to a McKinsey survey, at least a third of activities could be automated in about 60% of occupations. And in its recent Trends in Workflow Automation report, Salesforce found that 95% of IT leaders are prioritizing workflow automation, with 70% seeing the equivalent of more than four hours of savings per employee each week.
Switching repetitive tasks to RPA functions not only eliminates errors, it also garners significant cost savings. That’s because RPA addresses bottlenecks with workflows, data, and documentation while providing audit trails and reducing compliance expenses and risks. RPA can also boost legacy integration and record digitization and enable data-driven decisions and “path-to-cognitive” technologies, according to Technologent’s Kevin Buckley.
But as RPA expands to increasingly complex domains, the technology itself grows more complicated. This makes it harder for business decision-makers to understand where and when RPA might be appropriate, factoring in their industry and particular challenges.
RPA: What is it?
RPA is the category of software that automates tasks traditionally done by a human, using software robots that follow a set of rules and interact with enterprise systems via user interfaces. These robots can complete repeatable tasks, perform system integrations, and automate transactions from task-level to enterprise-level via scheduled orchestration.
There’s nuance within this definition, however. RPA often begins with what’s called backend task discovery, or process mining. An RPA client pulls log data from existing systems — including desktop, IT, and email apps and workflows — to identify root cause issues through recommendations, KPIs, and more. Task capture is the next step in the onboarding chain. It comes as employees move through a work process they’d like to automate by taking screenshots, using drag-and-drop designers, and pulling data like window names and descriptions together into a process definition document.
Most RPA platforms leverage AI to map tasks to automation opportunities and identify the most frequent patterns from the data, recording metrics from apps, including steps and execution time. Document understanding capabilities allow these platforms to ingest, analyze, and edit PDFs and images, even those with handwriting, checkboxes, signatures, rotated or skewed elements, and low resolutions.
Computer vision algorithms enable RPA software to recognize and interact with on-screen fields and components like Flash and Silverlight. Drawing on AI, optical character recognition, and approximate string matching, RPA robots can “see” virtual desktop interfaces via clients like Citrix, VMWare, Microsoft RDP, and VNC.
Types of robots
Not every RPA robot is created equal. Platforms such as UiPath offer three types: attended, unattended, and hybrid robots.
Attended robots act like a personal assistant that resides on a user’s computer to take a series of user-triggered actions and complete simple, repetitive tasks. By contrast, unattended robots require very little intervention to perform intensive data processing and data management workloads. Hybrid robots, as their name implies, are a combination of attended and unattended robots and deliver user support and backend processing in a single solution.
Choosing which type of RPA robot to deploy depends on the application. Because attended robots are tailored to the requirements of the user, they are a shoo-in for contact centers, field sales, retail, service engineers, and insurance agents. The scalable nature of unattended robots makes them a fit for application, claims, and invoice processing, as well as data and documentation search and retrieval. As for hybrid robots, they tend to work best in end-to-end scenarios like HR management, application processing, service delivery, and customer support and engagement.
Regardless of the bot type, RPA platforms typically leverage scalability to their technological advantage. For instance, startup WorkFusion claims its bots aggregate and share learnings across the bot ecosystem to create network effects from which all of its customers benefit.
Orchestration
RPA software lets customers manage up to thousands — or tens of thousands — of robots from a single dashboard. Customers can view the robots’ tasks and supporting documents, take remedial actions in the event of a bottleneck, and visualize automation complexity and payback costs. Some software offers toolsets developers can use to borrow prebuilt automation activities, integrate third-party components, and share and reuse components. RPA software also typically lets customers import their own machine learning models or choose from a marketplace of prebuilt options and keep tabs on versioning.
In the areas of AI and machine learning, Indico and other RPA providers apply techniques like transfer learning — where a model tailored to one task is used for another, related task — to deploy to unstructured content more effectively. The company’s out-of-the-box models, which were trained on large datasets of documents, ostensibly learn to analyze industry-specific data from just a few hundred training examples.
Connectors also add enormous value in the world of RPA. For example, RPA startup Bizagi integrates with Azure Cognitive Services to automatically recognize new kinds of paper forms and extract data from them. Sources include contracts, claims forms, emails, spreadsheets, purchase orders, and field reports. And Blue Prism offers a library that gives partners and customers the ability to create, share, and deploy plugins for the company’s RPA solutions.
Benefits
RPA can handle a vast number of different tasks, from contract audits and customer onboarding to commercial underwriting, financial document analysis, mortgage processing, billing form reviews, and insurance claims analysis. That is one reason the overall RPA market is expected to grow by more than 7% annually over the next few years to reach $379.87 million by 2027, up from $182.8 million in 2019.
Early in the pandemic, RPA companies like Automation Anywhere worked with health care centers to implement bots and automate laborious processes. For example, Olive, a Columbus-based health care automation startup, used a combination of computer vision and RPA to support COVID-19 testing operations by simplifying manual data entry. UiPath partnered with a Dublin-based hospital to process COVID-19 testing kits, enabling the hospital’s on-site lab to receive results in minutes and saving the nursing department three hours per day, on average.
Beyond health care, Gryps, an RPA startup focused on the construction industry, is applying machine learning to organize construction project files and documents. For the Javits Convention Center in New York, Gryps’ software automatically ingested over 20,000 documents and 100,000 data points, collated them, and handed them over to the Javits team, with estimates putting the savings at hundreds of hours of staff time.
The number of industries RPA touches continues to grow, with a Deloitte report predicting the technology will achieve “near universal adoption” within the next five years. According to the same report, 78% of organizations that have already implemented RPA — which see an average payback period of around 9 to 12 months — expect to “significantly” increase their investment in the technology over the next three years.
Security challenges
This isn’t to suggest that RPA is without challenges. The credentials enterprises grant to RPA technology are a potential access point for hackers. When dealing with hundreds to thousands of RPA robots with IDs connected to a network, each could become an attack vessel if companies fail to apply identity-centric security practices.
Part of the problem is that many RPA platforms don’t focus on solving security flaws. That’s because they’re optimized to increase productivity and because some security solutions are too costly to deploy and integrate with RPA.
Of course, the first step to solving the RPA security dilemma is recognizing that there is one. Realizing RPA workers have identities gives IT and security teams a head start when it comes to securing RPA technology prior to its implementation. Organizations can extend their identity and governance administration (IGA) to focus on the “why” behind a task, rather than the “how.” Through a strong IGA process, companies adopting RPA can implement a zero trust model to manage all identities — from human to machine and application.
A privileged access management (PAM) setup that can secure and govern RPA systems can also help. PAM systems allow enterprises to secure, control, and audit the credentials and privileges RPA technology uses without compromising the return on investment (ROI).
ROI
RPA challenges don’t stop at security. Deloitte reports that 17% of organizations face employee resistance when piloting RPA and that 63% of those organizations struggle to meet time-to-implement expectations. But the RPA’s return on investment often outweighs difficulties in deployment. According to the Everest Group, top performers earn nearly 4 times on their RPA investments, while other enterprises earn nearly double. And Gartner estimates that by 2024, organizations can lower operational costs 30% by combining automation technologies like RPA with redesigned operational processes.
“The first wave of robotic process automation brought the power of technology to users’ desktops in all industries and companies of all sizes. Today, we see a second wave emerging,” WorkFusion CEO Alex Lyashok recently told VentureBeat via email. “Cloud-based, AI-enabled robots [are] bringing intelligent automation to all enterprises.”
