In Short
USDM's robotic process automation (RPA) framework runs full circle: it defines the intended use of an automation, assesses its compliance, drives review and feedback through risk-based testing, and approves or adjusts it before release. Along the way, reusable components, open-source Java libraries, drag-and-drop ML tools, and custom activities let teams build and validate RPA faster while keeping a continuously compliant, GxP-ready tech stack.
The framework sets the stage. USDM’s robotic process automation (RPA) framework goes full circle from initiating the intended use of RPA and assessing its compliance, to review and feedback during risk-based testing, to approval or adjustments as it approaches its release into the wild. The multitude of tests within the framework has enabled USDM teams to gain valuable insight and improve the RPA process as it evolves.
In case you missed it, Part 1 of this blog series introduces the concept of RPA and how it can help your employees focus on their high-value work. Part 2 addresses RPA GxP-relevant use cases and provides a few examples.
RPA Components, Strategy, and Optimization
Reusable Components
Reusable components include application functions and data connectors, custom activities to replace default activities, developer tools, and workflow templates. These components contribute to RPA for simple tasks like checking or receiving emails, manipulating spreadsheets, and translating text, or more complex tasks like a management console configuration verification.
USDM reusable components are published as libraries that can be added as dependencies to automation processes. From there, they can be linked to component validation artifacts (for example, testing and requirements) that leverage previous validation.
Reusability is the key to efficient RPA development and validation. When a workflow is under design and development, existing requirements associated with reusable components and their associated test coverage can be leveraged for components that were already validated. When these artifacts exist in our cloud repository, we can easily generate reports for them.
Why reusability matters: Every component that carries its own validated requirements and test coverage shrinks the validation burden on the next automation. A risk-based computer software assurance (CSA) approach focuses testing effort where it changes patient safety or product quality, so reused, previously validated components don't have to be re-proven from scratch.
Open-Source Java Libraries
Open-source libraries for Java development have proven time and again that they provide reliable code to accomplish a variety of tasks like monitoring application logs, reading Excel files, and connecting databases.
The USDM RPA team leverages open-source Java libraries to automate complex web UI validation tasks and reporting. For example, we created a Java project using Selenium to interact with the DocuSign application and validate approximately 300 settings configured for the DocuSign account. This automated solution validates selected application settings to the value provided in the configuration specification document and includes items like checkboxes, radio buttons, select boxes, and input text fields. This automated solution also identifies other available settings and records them in the output report for impact assessment.
Configuration verification at this level is also where electronic-record and electronic-signature controls come into focus. Automations that touch GxP systems should be designed with 21 CFR Part 11 compliance in mind so that signatures, audit trails, and access controls remain intact end to end.
Reusability is the key to efficient RPA development and validation—and risk-based testing is what keeps that efficiency compliant.
Drag-and-Drop ML Tools
As artificial intelligence (AI) and machine learning (ML) become more commonplace—and you don’t need a developer specializing in AI to do all the work—ML models are easier to create and connect so that output from one model becomes input for the next.
As we work toward making RPA robots (bots) smarter and improving their cognitive processes, drag-and-drop ML tools will enable our developers to create intelligent workflows for our customers. As bots take on more cognitive work, clear oversight matters: pairing automation with AI governance and compliance keeps intelligent workflows accountable, traceable, and ready for regulatory scrutiny.
Custom Activities
An activity is a basic component of RPA. For example, your email program uses an activity to manage your messages according to specified rules, while your word processing application uses activities to format your documents.
Custom activities are created as we work through the USDM RPA Framework to define and build your product. Once these custom activities are complete, we can install and drag-and-drop them into workflows specially designed for your organization’s workflow processes.
In one case, USDM created a workflow to read Oracle unstructured data from various client configuration templates and put it into an Excel file with standardized column headers and rows, which allows automations to read the configuration settings. Standardizing data this way is also a data integrity safeguard—consistent structure makes records easier to verify, reconcile, and audit.
The USDM RPA Framework at a Glance
- Initiate & define intended use — establish what the automation is meant to do and why.
- Assess compliance — determine the GxP impact and the validation rigor required.
- Review & risk-based testing — test where risk is highest and gather feedback as the bot evolves.
- Approve or adjust — release validated automations, or refine them before they go into the wild.
- Maintain & reuse — publish components as libraries that carry their validation forward into the next workflow.
How USDM’s RPA Experience Can Help You
A cloud-based toolset to manage your RPA tasks is key. USDM’s Cloud Assurance is the right service for rapid implementation, validation, and maintenance to enable a continuously compliant tech stack (SaaS, PaaS, and IaaS) in your organization.
Our flexible cloud adoption model means that you can migrate your GxP workload to take advantage of built-in security and compliance while minimizing your cloud expenses and resources.
While you are able to focus on validating your business flows and exceptions, USDM will improve your processes and save you time and money by using validated RPA applications for GxP, including repeatable configuration and intended use verifications for your cloud system. To keep that validated state durable over time, many teams adopt a disciplined approach to validation lifecycle management so automations stay compliant as systems change.
FAQ: RPA Validation and the USDM Framework
What is the USDM RPA framework?
It is a full-circle approach that moves an automation from defining its intended use and assessing compliance, through review and risk-based testing, to approval or adjustment before release. Continuous feedback during testing lets USDM teams improve the RPA process as it evolves.
How does reusability speed up RPA development and validation?
USDM publishes reusable components as libraries that can be added as dependencies to automation processes and linked to existing validation artifacts. When a new workflow uses a component that was already validated, its requirements and test coverage can be leveraged instead of being re-created, which makes both development and validation more efficient.
How are open-source Java libraries used in RPA?
USDM uses open-source Java libraries for tasks like monitoring application logs, reading Excel files, and connecting databases. In one example, the team built a Selenium project to interact with DocuSign and validate roughly 300 account settings against a configuration specification, recording other available settings for impact assessment.
How does RPA stay compliant in a GxP environment?
Automations are built and tested within a risk-based framework, with attention to electronic-record and electronic-signature controls and to keeping data consistent and auditable. Cloud Assurance then provides ongoing implementation, validation, and maintenance to keep the tech stack continuously compliant.
Where do machine learning tools fit into RPA?
Drag-and-drop ML tools let developers connect models so one model's output becomes the next model's input, helping make bots smarter and more cognitive. As automations take on more decision-making, governance and oversight keep those intelligent workflows accountable and traceable.
Ready to accelerate your RPA journey?
Contact USDM today to get started with or accelerate your RPA uses—and to build validated, GxP-ready automations on a continuously compliant foundation.
On-demand Webinar
Here’s a sneak peek of the information presented by Stepheni Norton and Jim Lyle in the How to Maximize Your GxP Use of the Public Cloud webinar. To learn more about our internal framework for RPA and potential GxP use cases, watch the full-length on-demand webinar.
Watch the full-length on-demand webinar How to Maximize Your GxP Use of the Public Cloud.