RPA is a non-invasive approach to automating existing processes without the need to upgrade your entire technology platforms – creating the basis for even more intelligent automation that frees up your human workers to apply their brainpower to the things people do best.
RPA allows business and IT to collaboratively automate processes across the enterprise.
Robotic process automation (RPA) is a technology that executes manual and repetitive processes efficiently by mimicking what users do on their computer. This can include checking emails, searching for and extracting information from web pages, entering data into ERP systems, performing basic calculations in an Excel spreadsheet, or saving results on a SharePoint site. The main advantage of RPA is its ability to integrate and orchestrate many different activities using different systems to automate business processes – very cost-efficiently.
RPA integrates traditional legacy systems, but also creates a platform for intelligent automation using advanced AI and ML technology.
If you are a large organisation, RPA is a powerful and cost-efficient step along your digital transformation journey. It enables you to re-use, integrate and orchestrate existing systems instead of developing expensive new technologies that will disrupt your business and overburden your staff. An RPA robot or ‘bot’ can be developed easily at low cost and updated flexibly during process changes, and uses the same technologies that users are already familiar with.
Besides working with traditional legacy systems, RPA enables more advanced artificial intelligence (AI) solutions such as machine learning (ML) systems. RPA robots or ‘bots’ are a cheap way of extracting the data that these solutions need, as well as orchestrating the execution of actions recommended by AI/ML algorithms in downstream enterprise systems. RPA can also integrate advanced natural language processing (NLP) modules to enable advanced text extraction and processing activities. All these things are crucial building blocks for organisations wanting to automate intelligently and ultimately extract the maximum value from their human and data resources.
Applied intelligently as part of a coordinated approach, RPA is a straightforward way to power enterprise-wide digital transformation.
RPA allows business and operational units to:
- automate back-end, repetitive, manual and rules-based processes
- capture analytical and valuable process data that can be used for AI (especially machine learning) components
- shift the focus of human talent from time-consuming, low-value tasks to high-value, innovative and strategic work.
By combining enterprise-grade RPA platforms such as Blue Prism with AI, ML and NLP, you can create a foundation for intelligent automation. The ultimate vision of intelligent automation is to create collaborative digital workers to empower the human operators they work alongside. The idea isn’t to replace humans with digital workers, but to enable digital workers to become an intrinsic part of the fabric of a future workplace.
Implementing robotic process automation needn’t be complicated. It works particularly effectively if an organisation sets up an RPA centre of excellence that delegates process-building and software robot management to operational staff (who have the most knowledge about optimal process workflow). It also helps to know the common challenges in RPA implementations and the best-practice solutions for overcoming them.
Common challenges in RPA implementation
1. Top-down vs bottom-up opportunity identification and governance
In a top-down scenario, an organisation centrally selects automation opportunities that are in scope with a feasibility and cost-benefit study. The process teams follow the decision. In a bottom-up approach, automation opportunities are identified in a federated fashion—where the process teams’ input is welcome and considered when choosing the process(es) in question. Both options have their merits and challenges:
Leads to an automation scope with the best cost-cutting potential, thanks to the central feasibility study
More efficient, better overall view: the RPA team gets more efficient at achieving the overall automation goal because the process is larger, follows a long-term vision, is planned centrally, and produces more re-usable scripts
Risk of a low adoption rate because process teams resist decisions taken by management
More likely to get the support of process teams, as the project scope is initiated by them
Since the feasibility study is less relevant, the scope may bring only minimal cost-saving -opportunities
Less good at overcoming silos because synergies between automation opportunities aren’t considered
Given the various pros and cons, we often see a combination of top-down and bottom-up approaches. At many organisations process teams are involved in the decision making process, but don’t dictate the final decision. Key strategic areas of business for automation are defined from the top down, and minimum acceptance criteria are communicated up-front to ensure all opportunities identified by the process teams have certain cost-savings potential to qualify for automation.
Identifying processes isn’t the only consideration. From a firm-wide automation programme governance perspective, it’s important for organisations to define an IT strategy comprising areas for automation so that they can centrally invest in key areas and technologies in alignment with a common organisational vision. This will enable optimum integration and maximise the return on investment (ROI) and performance benefits. Blue Prism’s best practice guidance for implementing a robotic operating model (ROM), for instance, recommends defining a structure for RPA governance including executive accountability, roles/responsibilities, planning, management, and oversight.
2. Change management
Part of change management is understanding how RPA works and how and where problems can arise. Incidents can occur when software robots are used, but in most cases they’re due to unexpected changes in underlying third-party applications (the systems the bots are working on) or system performance rather than the robotic process automation itself. RPA software such as Blue Prism is rule-driven, with the ability to respond to and interact with applications and systems, and is deterministic. Often, RPA users only detect underlying system changes after a robot runs when an exception was produced. This can result in the robot being down for some time, as well as the risk of inaccurately outputting information that is gathered by the robot.
RPA robots are a new type of ‘user’ for most IT departments and business units, which means that robots should be included in the existing development and testing framework. For example, it’s advisable for every process owner to formally ask the IT team to include the robot in the system-level user acceptance test (UAT) plan to identify system issues when implementing RPA in production. This helps prevent robot-related incidents. In terms of implementation it’s very useful if the subject matter expert draws up a process definition document (PDD) outlining the goals, input, output and steps of the process.
Address these challenges with best-practice responses and you’re well on the way to a smooth RPA implementation – an important step in your journey to intelligent automation.