Where do you start thinking about how to define an RPA operating model?

Editor's note: This post introduces an RPA operating model. For a more in-depth look at how to successfully create an automation operating model, be sure to check out Benefits of an Automation Operating Model (Part 1).

There’s no one-size-fits-all approach when it comes to Robotic Process Automation (RPA) implementation. Each automation rollout needs to be adjusted to a company’s organizational setting and requires thorough planning: How can the engagement of key stakeholders be ensured? What should the scope, roadmap, and delivery approach include? From where will the necessary funding be secured? Another essential element in the bigger picture of preparing an RPA deployment involves designing an operating model. The earlier an organization-wide operating model is defined, the higher the likelihood of RPA’s benefits to be realized ongoingly and long-term.

Being proactive in delineating an RPA operating model involves the translation of a strategic automation plan into operating requirements and decisions. Such a “blueprint” dictates what processes will be automated, to which employees RPA responsibilities should be assigned, how the software will be integrated into the IT landscape, and more.

Establish the right in-house skillset

Every RPA implementation team must have a deep understanding of the processes for automation and the skills required to develop an automation program. A lack of adequately trained or appropriately skilled automation specialists can risk the success of an RPA implementation project and minimize opportunities for achieving maximum benefits. Varying degrees of responsibility will be delegated to process subject matter experts, scrum masters, process analysts, solution architects, developers, controllers, as well as IT security and infrastructure specialists. This kind of structure and organization within the team will help to define the accountabilities of the involved individuals and enable the adequate flow of communication.

Determine the level and structure of operational accountability

A primary consideration when designing an RPA operating model is determining the extent to which the RPA unit will have operational responsibilities. Will the RPA unit be a robust operating center that directly manages the digital workforce, or will it assume an advisory role that delivers oversight of the automation program? The RPA organization can also operate with different levels of centralization: centralized, decentralized, and hub-and-spoke. These varying degrees of centralization may be adjusted to the resources available as well as how the company implementing RPA is itself structured, whether there is firm, corporate control from centralized headquarters or autonomy is granted to individual offices or departments. A centralized operational model gives authority to a Robotics Operation Center for ensuring the smooth functioning of RPA, while a more decentralized model relies on the business and IT departments to select the processes for and carry out automation.

Ensure strong governance procedures

RPA deployments can easily slump or even fail without preemptive planning, and adopting RPA in an unorganized, complacent fashion can lead to some challenges. Establishing a well-developed RPA governance strategy must be a part of the journey from the early stages and should extend far beyond just the initial implementation phases of RPA. Maintaining comprehensive governance practices include consistently communicating between business and IT representatives, complying with regulatory requirements, laying out operational objectives, as well as having a plan for a potential fall-out of RPA. The designation of an RPA organization leader and the establishment of an automation council to oversee the alignment of the automation project at a company-wide level is also best practice for creating the foundation for a successful RPA initiative.

Define key operational processes

Streamlining the functioning of an RPA program requires several vital operations to be designed: identifying processes, prioritizing and deploying them for automation, as well as developing and maintaining the automated workflows. Documenting these processes should involve creating detailed user stories, defining business requirements, as well as testing, managing, and reporting. It is important to note that these functions depend on a company’s priorities, resources, and needs. For example, deciding which processes to automate first may depend on whether the automation project focuses on cost reduction or error elimination.

Track KPIs

Automation enables the ability to reduce operational costs while maximizing productivity, therefore achieving financial returns within less than a year is achievable. Unexpected operational costs or a downturn in sales can, nonetheless, limit such success and put producing returns at risk. To manage the performance of RPA within an organization, a performance management framework – that defines process SLAs as well as financial, employee, and virtual workforce KPIs – should be established. Continuous evaluations of these benchmarks are essential to ensure that RPA will eventually deliver its benefits on time and within budget.

It all comes down to avoiding long-term risk

Each company will need to fulfill unique requirements when starting their RPA. Nonetheless, mitigating the risks associated with RPA implementation represent a commonality among all adopting organizations. Determining an RPA operating model is one of the first steps to creating a foundation for successful and sustainable automation. Only with a well-designed, comprehensive framework for deploying RPA will you be able to create lean, efficient, and cost-effective business operations not only fast, but also long-lasting.

Don't get stuck. Learn how to scale your RPA, step by step.


This is a companion discussion topic for the original entry at https://www.uipath.com/blog/rpa/rpa-operating-model

I am very much new to this UiPath Forum, just have a query that need a simple reply "How would I make a model that can estimate time of Human base tasks and a Robotic one.