Sometimes referred to as “smart automation” or “intelligent automation,” Gartner defines robotic process automation or RPA as an umbrella term for advanced software systems that can be programmed to perform a series of tasks that previously required human intervention. Other robotic solutions incorporate machine learning and include cognitive computing and artificial intelligence.
Often referred to as "bots", these “digital FTEs” or “human proxies” work and interact with existing systems as a person would. RPA vendors claim RPA software usually costs about a third of an offshore employee or one-fifth of an onshore one with the additional difference of being able to perform work nonstop and with no human errors – IF programmed correctly.
The days of weaving a cacophony of software, hardware and other technologies to try and solve a problem, while still being practised, have become less critical. Matthew Thomson, SVP and GM for APJ at Kofax, says an ecosystem that provides ease of use, frictionless buying, security, interoperability, and a library of third-party add-ons wins out.
"RPA has matured to take its place as a critical enterprise automation fabric technology. RPA products and supporting ecosystems have evolved significantly. Enterprises trying to build stable, scaled RPA programs grapple with the challenges of RPA’s past but also look to its potential to help deliver the autonomous enterprises of the future."
Leslie Joseph
The hiccups of RPA
This ability to work nonstop is an attractive proposition for businesses. However, the ability to scale the technology, to move beyond the few repetitive tasks that were accomplished in the pilot stage, and to reap the benefits of economies of scale remained one of the early challenges of RPA.
In the 2022 IDC APJ automation survey commissioned by UiPath, over 9 in 10 organisations in Asia Pacific and Japan (APJ) do not have an enterprise-wide RPA deployment. The report cited the lack of skilled talent (59%), particularly automation skills, as hampering enterprise-wide scaling of automation.
Aside from the skills issue Kofax's Thomson cited resistance to change as a cultural norm making it difficult to get buy-in from stakeholders. He also noted that many companies have legacy systems that are not well-suited for automation. In addition, data quality and compliance requirements can also have an impact.
Leslie Joseph, a principal analyst at Forrester, clarifies that scaling of RPA is not just an Asia problem, adding: "Our research has consistently shown that there are three classic impediments to RPA scale that manifest at different levels of program tenure and maturity."
He explains that among early-stage programs, the challenge lies in establishing a repeatable approach to process discovery and assessment. For those in mid-stage programs, it is establishing a robust automation governance model.
"The third, usually manifesting in long-running and mature programs, is bringing about all-around culture and workforce transformation by fostering an automation-first mindset," elaborated Joseph.
Overcoming the hiccups
According to Jess O’Reilly, area vice president for Asia at UiPath, automation across the enterprise requires business leaders to think about a holistic automation strategy. They need to recognise that they cannot implement automation technologies in isolation.
"They need to give employees an active role in the automation initiative and leverage low-/no-code solutions to allow non-technical users to build and deploy automation applications to support their day-to-day work," she continued.
To these, Thomson suggested using a coherent RPA platform, avoiding overly complex processes, and implementing a governance framework. Having a clear roadmap also helps, as does measuring and monitoring results.
Joseph advised that scaling must go hand-in-hand with value. "Organisations that manage to scale their RPA programs also institute robust mechanisms for value tracking and reporting across 'hard saves’ as well as intangible benefits," he continued.
The promise of smarter robots
UiPath's O'Reilly noted that the development of advanced software robots, empowered with recent AI innovations, marks a vital transformation to the future of automation. AI expands the range of processes that businesses can automate today.
"As we look ahead, AI will increasingly become the cornerstone of the automated enterprise, and AI-powered automation is expected to be a critical requirement for business excellence, customer experience and competitive success," she added.
More recently, the buzz around generative AI (GenAI) has raised expectations as to what the technology can do for enterprises. Bardia Eshghi, an industry analyst at AIMultiple, opined that business leaders should treat GenAI and RPA as transformative technologies capable of reshaping their companies’ operational efficiency.
Acknowledging its potential to improve the efficiency of RPA bots, Thomson explained that GenAI could be used to generate automated workflows and process models, which could then be used to train RPA bots.
"Overall, I believe that GenAI has the potential to make RPA more powerful and versatile. This could lead to increased adoption of RPA by businesses of all sizes," he continued.
For O'Reilly specialised AI presents distinct advantages in enterprise AI, compared to large foundational models trained on general knowledge.
"Specialised AI focuses on specific tasks and once trained on task-specific data, can be optimised for specific use cases within the organisation, translating to fast, accurate and tailored solutions that are not only cost-effective but which also deliver high-value outcomes."
Jess O'Reilly
Specific to RPA, Forrester sees GenAI being used either to improve the developer experience or to enhance cognitive abilities when dealing with natural language or unstructured data.
Gartner says ultimately the only limit to RPA’s scalability is a company’s ability to keep pace with the change management needed to sustain it.
Lessons from past RPA implementations
The Forrester report, The 10 Golden Rules of RPA Success" outlines ten rules for organisations looking to get a head start in their RPA initiative. These include weaving RPA into the company's automation fabric, building a sustainable business value model for RPA, and reporting value in ways that matter to business leadership.
To ensure RPA deployment success, Thomson lists out several action items: to start small and scale gradually, get buy-in from all stakeholders, focus on the right processes meaning repetitive, rule-based and high-volume activities, create a clear roadmap, measure and monitor, invest in training and establishing a governance framework.
"Most of all, be patient! RPA is a long-term investment that will garner the results you are looking for, but you can’t expect to see results overnight. Aligning with the right vendor is also helpful in attaining your goals. Finding a partner with a low-code or no-code platform is important along with expertise to help avoid common pitfalls and to get the most out of your RPA investment."
Matthew Thomson
For her part, O'Reilly commented that harnessing the full potential of automation in the enterprise necessitates collaboration.
She cautioned that automation teams should not operate in silos; they need to constantly engage with key stakeholders from other departments and functions to ensure that automation initiatives align with broader business goals.
"It is also important to move past a piecemeal approach and focus on establishing a strong foundation for automation in the organisation," she suggested.