Doug Wendler, chief executive of Machina Automation, explains the increasing popularity and uptake of robotic process automation (RPA), focusing on how it has recently evolved and how it is expected to continue this evolution, and whether RPA’s automation of inefficient processes is a hindrance or an opportunity
RPA is becoming increasingly popular with energy firms. Proponents say it brings the benefits of automation – such as the elimination of error and the freeing-up of human resources – relatively quickly and with minimal disruption. This is because the software sits on top of existing IT applications, often communicating across multiple systems, which can also increase process efficiency.
Some IT experts worry RPA is being used to automate inefficient processes, thereby perpetuating them rather than addressing any underlying inefficiency. Others see RPA as an essential step on the journey towards full digitisation. So far, its use in energy firms is largely confined to back-office and logistics processes. However, Doug Wendler, chief executive of Machina Automation – a technology firm specialising in RPA for the energy industry – expects its use to expand significantly in the coming years.
How advanced is the energy industry in the adoption of RPA? Have any trends emerged?
Doug Wendler: It is still early days. Probably less than 50% of companies have formal RPA programmes and, of those, only relatively few have moved beyond pilot or proof of concept into an enterprise‑wide programme aimed at driving return on investment. We expect more companies to move into enterprise programmes this year.
Currently, the majority of programmes are aimed at automating only the most straightforward processes, such as monthly reporting or reconciliation. For example, we have one client that was reconciling more than 50 accounts across multiple banks to their financial close system. They were able to reduce the time required to complete the reconciliation by over 80% while improving accuracy.
Within the energy industry there is a wide divergence in RPA use cases, but there is a trend emerging around volumetric reconciliation. RPA can be used to log into counterparty and regulatory systems and check that everyone has the same volume data at every step of the trade cycle. As well as increasing efficiency, this could help firms ensure, for example, that their volumetric capacity has not exceeded pipeline capacity, or breached a regulatory threshold that would incur a fine.
Is there a risk that RPA could harm an organisation by perpetuating suboptimal business processes?
Doug Wendler: While it is best to optimise a process prior to automating it, there are benefits to automating even an inefficient process. It will provide you with better data with which to understand how it can be improved. Machina Automation helps firms identify which processes would benefit most from automation, often looking at what adjustments need to be made to a process to help increase automation efficiencies.
Is RPA a tactical rather than strategic solution?
Doug Wendler: RPA can certainly be deployed tactically, but we believe it also has a key role to play in a firm’s strategic digitisation agenda. Simple automation of manual tasks is a great use case for RPA and can be temporary as a highly efficient method to expedite a one-time or short-term activity. For example, during a merger or acquisition when systems are being integrated, it can be more cost-effective to use robots than to do the integration manually. But either way the project is a temporary one.
We are also seeing growing appetite in the industry for ‘citizen developers’ – or a robot on every desktop. Organisations are contemplating making RPA accessible for individuals to self-automate their everyday activities. If achieved, this form of automation will often be tactical and short-lived, but very efficient in reducing an organisation’s overall manual effort.
However, RPA can also allow firms to carry out processes at an enterprise level, rather than an individual systems level, and this will be the next move. For example, using bots can produce a co-ordinated end-of-day risk management process rather than automating several different end-of-day runs.
Furthermore, RPA is a critical building block in a longer-term automation lifecycle. In a five- to 10-year evolution, RPA is a necessary first step, and the outputs of RPA – the digital exhaust – will help enable more sophisticated process-related intelligent automation. Companies that do not develop an RPA strategy won’t have the necessary process-level data for certain process related artificial intelligence (AI) development in the future.
How do you expect RPA to evolve?
Doug Wendler: More RPA and other specialty tools are now being developed and utilised to identify the best use cases for RPA and to enable more efficient implementation. The timing of this development is great for energy firms because they will be able to take advantage of these tools from the start of their enterprise programmes rather than trying to incorporate them after broad RPA adoption – as is the case in some other industries.
Tools such as UiPath are just one of the building blocks necessary for a holistic digital automation strategy. Other key tools include process mining, workforce orchestration and cognitive tools. The landscape is still emerging, and the full picture of how to best use these tools is still developing.
In the medium term, we expect an increased adoption of machine learning and other AI tools. RPA, as we know it today, is constrained by rules programmed in when automations are developed and must be updated to remain current. The addition of machine learning capabilities enable automations to learn from the decisions made by humans and from the data with which they interact. This allows the automations to grow the rule set they operate by, becoming ‘smarter’ and thereby reducing exception cases and improving efficiencies. The utilisation of machine learning is an extension of RPA and not a departure from it.
What should energy industry participants be doing now?
Doug Wendler: Those who haven’t yet investigated RPA should do so quickly. Those who have begun a programme of RPA work should investigate opportunities beyond the back office and look towards complementary orchestration and cognitive tools to widen the scope and improve the benefits of their RPA use. RPA can play a significant role at various levels of complexity, and the sooner organisations capitalise on it the better.