Wiki: Robotic Process Automation

Robotic process automation (RPA) is used to automate manual, labor-intensive, time-consuming, rule-based, or repetitive tasks. Typically, a bot performs these tasks more efficiently than humans, while also being available 24/7 and easily interacting with in-house applications, websites, and user portals. They work well in a static environment and can log into different applications, copy and enter data, open emails and attachments, carry out calculations, and much more.

RPA is used most effectively in cases where:

• IT resources and budgets are limited
• Organizations are working with backend applications that lack APIs and legacy IT systems

But, before deploying RPA, you must understand its benefits and limitations, how to optimize its evolving capabilities, and the value it provides.

In an IT environment, most business processes are dependent on multiple IT systems that don’t interact with each other. With RPA-based digital transformation, organizations can make these processes faster, more accurate, and efficient– freeing up human workers to focus on tasks that require emotional intelligence, reasoning, judgment, and person-to-person interaction with customers.

There are two main types of RPA:

• Attended RPA – Robots act as personal assistants on users’ computers, performing a series of user-triggered actions to complete simple, repetitive tasks and streamline the workflow
• Unattended RPA – Robots require little to no human intervention while executing complex functions, such as intensive data processing, data management, or even complete back-office functions

Over the last few years, a number of findings, about broadly applying RPA to broken processes, have emerged based on data from early-stage RPA deployments:

1. 30-50% of initial RPA projects fail
2. Only 3% of bot deployments reach their intended scale
3. 63% of business leaders cite unsatisfactory implementation speeds

RPA offers value to every enterprise on an automation journey, but the challenge is determining how to deploy it appropriately. While RPA delivers tactical value, it can often distract from a strategic transformation. Focusing only on RPA can perpetuate problems associated with legacy systems—allowing issues to go unattended and failing to streamline processes that would ensure superior customer experiences.

If an enterprise is suffering from siloed operations and bottlenecks, an army of robots performing repetitive functions isn’t necessarily the right fix. While RPA allows for the automation of existing processes, it fails to address the transformational needs of an enterprise—reengineering processes and incorporating digital technologies. A McKinsey report on the risks of RPA suggests that, “Taking an end-to-end view of the outcome […] is better than applying a robotic band-aid to a particular pain point.”

• RPA automation is non-invasive in nature, which means it is most useful in streamlining operations centered around clerical activities, carried out using a collection of legacy applications to perform “swivel chair” tasks. It is also useful in cases where you want to achieve results quickly, without going through the hassle of trying to re-engineer existing legacy systems

Instead, RPA is more effective when combined with intelligent process automation (BPM), especially in cases where:

• There is a need to automate the flow of work across different tasks
• The tasks or processes being considered for automation require significant human intervention and domain expertise for completion
• Task completion involves decision making based on sophisticated, multi-step business rules
• Business requirements or technology environment keep changing; traditional RPA-based implementations can be fragile and highly sensitive to such shifts
• RPA implementations fail and humans need to correct the issues, thus negating any gains from automation

Even though at first RPA alone might appear to be the most suitable solution, it is important to treat RPA technology as one part of a holistic automation approach.

Innovation leaders are responsible for ensuring that their organizations are on top of technology trends. This means that they likely have either implemented RPA already and are trying to scale up their investment, or they are aggressively exploring and planning to adopt RPA. At this point, most large organizations have shifted their focus toward scaling their automation efforts across their whole enterprise.

Like many other modern technologies, RPA continues to evolve at a rapid pace. In fact, Gartner has coined the term “hyperautomation” to describe the next stage of RPA evolution, including it in its list of the top 10 strategic trends for 2021.

The idea behind hyperautomation is to establish an integrated platform that enables organizations to automate multiple business processes quickly and intelligently. For organizations to effectively scale more than just simple and static processes, they must employ an ecosystem of tools to discover, automate, optimize, and monitor the complex processes that can deliver a significant ROI.

A Gartner report, titled “Move Beyond RPA to Deliver Hyperautomation,” suggests that no single tool can entirely replace the human workforce. Hyperautomation is comprised of RPA and BPM, combined with the goal of increasing and improving AI-driven decision making. Implementation of these technologies will likely provide significant business opportunities by providing real-time insights—enabling organizations to visualize how functions, processes, and key performance indicators interact to drive value.

Furthermore, Forrester states that in the coming years, the market will shift from implementing standalone RPA to a vendor-agnostic platform that supports multi-vendor RPA implementations.

Before kicking off your RPA journey, it’s important that you outline your automation plan by:

• Establishing clarity on why you are implementing RPA – You must have concrete reasons for picking RPA over the myriad other technologies available. Think, “why RPA?” to identify your specific objectives and criteria for success

Are you looking for reduced FTEs, improved task-level productivity, faster end-to-end process cycle times, or something else? Oftentimes, a broader process management need is disguised in the form of lesser issues, such as task-level productivity

• Understanding if broken processes are the real problem – RPA is not meant to fix broken processes or bridge silos across an organization. If you are struggling with disjointed or mis-managed processes, you may want to consider BPM

No matter how well you optimize and automate the human-performed activities within disjointed processes, you cannot achieve your desired outcomes without streaming the overall process. Essentially, your process has to be well-defined, structured, and mature for RPA to be an effective solution

• Identifying the right processes – At first glance, automation is warranted for any set of structured, error-prone, repetitive, and high-volume tasks. However, not all processes are created equal. Even if certain activities within a process are repetitive and structured, if they are liable to change very often, the effectiveness of RPA would be nullified
• Measuring, simulating, and benchmarking the process – You must be able to measure key process parameters, such as overall cycle time, throughput, and task-level attributes, and then gauge how they impact business performance. At the same time, you must simulate each process with various “what if” scenarios and assess if the proposed RPA solution would really impact business performance to the degree predicted
• Accounting for exceptions in the process – You must support your RPA implementation with a BPM platform. This provides a process-level, exception-handling mechanism that can seamlessly define a fallback option in the form of human intervention or through rule management. Operationally, it is crucial to enable exception handling through a process management solution before RPA can be implemented effectively
• Establishing visibility to facilitate administration of process and bots – Scalability requires robust and effective monitoring of processes and bots. Use process-level monitoring to track what’s happening in the system, including human activities, bot statistics, exceptions, queue monitoring, process-level and queue-level alerts, and overall process statistics
• Weighing the requirements from auditing, security, and compliance – Leverage a BPM-based RPA solution to ensure that all audit trails (including user-level, bot-level, and item-level audit trails) are available through a single system. From a security standpoint, it is imperative that you prioritize monitoring bots
• Considering horizontal and vertical scalability – The real value of RPA lies in vertical scaling to meet increased volumes through bot cloning. Introducing a BPM platform, that can manage the process at an elevated level, further provides options to realign and rearrange the process by splitting it and adding a variety of bots across parallel steps
• Exploring platform-based RPA for enterprise-grade automation – Different use cases call for different types of bots—where basic digitization requires relatively simple bots that rely on a record-script-playback mechanism, enhanced digitization requires rule management and data extraction capabilities. Platform-based RPA solutions can provide all these functionalities, alongside the ability to include various types of bots – simple, complex, and cognitive.
• Understanding the prerequisites for effective RPA – RPA is only part of a solution, and works best when paired with other capabilities, such as AI and BPM. For example, adding bots on top of a rigid and broken, ERP-based process would likely not yield the desired results. In fact, RPA alone can only be applied to the most structured and static parts of any process. While implementing RPA alone might successfully highlight existing rigidity in your processes, it could end up reinforcing it further through inflexible and granular automation

To learn more, download our eBook: Getting RPA Right.

RPA works best in situations where application interfaces are static, processes don’t change, and data formats remain stable—criteria that are increasingly rare, especially altogether, in such dynamic, digital times. This is where BPM, paired with RPA, can fill the gaps.

• RPA and BPM can be combined into a powerful solution for businesses that are struggling to boost the effectiveness and productivity of their business processes. RPA exacerbates the value achieved using a traditional BPM system
• BPM-enabled RPA allows businesses to drive continuous process improvement and catalyze their digital journey. Organizations can gain complete visibility across processes—including human activities, bot statistics, exceptions, queue monitoring, and process- and queue-level alerts—and then balance the workload between bots and humans

Image Credits:  Craig Le Clair, Vice President and Principal Analyst, Forrester

Benefits of BPM:

• Seamless management of customer-facing processes and applications
• Better execution of structured, unstructured, and adaptive business processes
• Streamlined coordination of work between people, systems, and processes
• Creation of an easy-to-use user interface (UI) for humans
• Robust API integration with third-party applications
• Transformation of business processes through automation
• Improved end-to-end customer experience through connected user interfaces and processes

Benefits of RPA:

• Easy identification of areas for improvement and access to process insights
• Automation of high-volume, repeatable tasks within existing business processes
• Quick mimicry of behaviors and patterns, especially the way workers interact with applications
• Automation of single-user tasks involving multiple applications
• Integration with backend, legacy, and new systems using a UI wherever APIs are not available

Invoice Processing in BPM:

Limitations:

• Valuable time spent on repetitive and mundane tasks
• Manual data entry leads to increased errors
• Longer cycle times to process invoices
• Complexity in integration with different shadow IT artifacts and third-party applications

RPA complements invoice processing in BPM:

Advantages:

• Increased process intelligence to identify RPA opportunities
• Automated mundane and repetitive tasks
• Faster invoice processing using robots that can work 24×7
• Decreased errors and improved straight-through processing
• Seamless integration with third-party applications
• Human workers are freed up to just perform managerial approvals and exception handling

ATM dispute resolution process execution with standalone RPA:

Limitations:

• Weak exception management
• Siloed bot execution
• Subpar process orchestration
• Poor data visualization

BPM provides a centralized platform for bots and humans:

Advantages:

• End-to-end process orchestration
• Seamless human and bot collaboration
• Effective exception management
• Better data visualization
• Improved audit trails
• Scalable RPA

• Agile RPA

      BPM strengthens RPA and bridges siloed processes

• Exception Management

      All business exceptions are automatically routed to BPM, allowing critical activities to be handled by human workers

• Scalability

      BPM-enabled RPA offers an endless scope for automation, especially enabling the reuse of bots across processes

• Compliance

BPM provides seamless audit trails and secure automation

• Data Visualization

BPM enhances RPA with process insights to enable end-to-end process orchestration and identify bottlenecks

• Efficient Governance

A vendor-agnostic automation platform, featuring dashboards and bot management, support BPM and RPA

• Better ROI

BPM makes RPA more scalable across various processes, without sacrificing on agility, thereby improving and increasing ROI

Robotic Process Automation (RPA)

• Involves automating rule-based, repetitive tasks to make processes less labor-intensive for humans
• Works best with static business processes that involve structured data—where the process is the same every time and data is sorted into well-defined fields, as in a spreadsheet
• Can only work if decisions are made using “if/then” statements. The process being automated cannot require any human judgment
• Cannot make judgments about data or learn from experience to improve processes, unlike machine learning applications
• Cannot handle unstructured data, such as text and images. It needs to know exactly what data to look for and where to find it

Hyperautomation

• Is an emerging set of new technologies that combines fundamental process redesign with RPA and machine learning (Hyperautomation = RPA + BPM + Process Mining + AI and ML)
• Involves cognitive abilities (AI and ML) and is suited to handle business processes involving text and images
• Builds on the concept of transfer learning, where a model trained to perform one task is used for another, related task
• Can translate unstructured content into structured data so it can be plugged back into an RPA platform
• Can mimic behaviors, learn from experience over time, and make judgements

Business leaders and decision makers must look for a robust RPA system that provides an integrated repository and enables reporting, as well as orchestration capabilities for human and digital workers.

Questions to Ask While Identifying the Ideal Solution for Your RPA Journey:

Strategy Perspective

• Will the solution help improve overall business results and reduce operational costs?
• Will the solution help ensure business continuity during uncertain times?
• Will the solution enable significant process improvements and enhance the overall customer experience?

Business Perspective

• Will the solution enable troubleshooting, auditing, and regulatory compliance?
• Will the solution help reduce TATs?
• Can the solution help identify processes that eligible for an RPA implementation?
• Is there a fallback plan in place if the implemented bots malfunction or if a business exception arises? How is this guaranteed?
• Is the solution navigable and easy-to-use for business users?

IT Perspective

• Does the solution provide a platform that supports scalability?
• Can the solution enable a smooth shift from standalone RPA to holistic automation?
• Is the solution compatible with the underlying architecture and infrastructure of the company?
• Will the solution integrate well with emerging technologies implemented in the future?

To learn more, download our eBook: Selecting the Right Digital Partner for Your RPA Journey