Robotic Process Automation Already Impacting Insurance

If there’s any industry that’s ripe for robotic process automation, it’s insurance, with all its paperwork, verifications and checks and balances.

The term -- or its shorthand, RPA -- burst into the lexicon fairly recently, offering a new, leaner and smarter way to manage the countless mundane, routine tasks that gobble up organizational time and resources. That’s how it’s being pitched, at least.

It’s definitely a growing space. Recent data from Transparency Market Research estimates that the IT robotic automation market will expand at an annual rate of 47%5 over the period between 2016 and 2024, rising to a valuation of $5 billion within the next three years.

Nigel Walsh, a partner at Deloitte, recently examined the impact of RPA on the insurance industry, noting that its automation capabilities “enable more focus and time spent on the customer, and can mean better experiences not only for customers but for employees and agents as well.”

Examples of insurance processes being automated through RPA include first notice of loss, fraud checking and policy renewal, including data gathering and recalculating policy premiums, Walsh says.

It’s early in the RPA journey, so case studies are still few and far between. But many feel the current generation of digital solutions isn’t enough to address insurance company administrative bottlenecks. A pitch from UiPath, one of the leading RPA vendors, sums it up this way:

“Many insurers already implement some level of automation, be it basic automation like scanning PDF documents or more advanced automation of entire workflows. However, these solutions are disparate and come with their own set of challenges that can further add to the complexity and difficulty of processing. The quality of output from these automations depends largely on the quality of programming. Every time there is a change in the process, the entire programming needs to be overhauled. Another shortcoming is that they still rely on manual input from human workers to process and navigate data between different systems, which burdens the claims process, leaves it exposed to errors, and adds extra cost. This is where robotic process automation comes in as a fully integrated, end-to-end automation solution that overcomes the abovementioned obstacles. RPA works at the presentation layer, integrating with all applications and systems, including legacy systems, without creating disruption. It is entirely scalable according to necessities and has up to 100% robotic accuracy.”

Wipro recently posted a case study involving an unnamed US health insurer that saw substantial benefits from RPA. As Wipro describes it, healthcare payers process thousands of contracts each year, which need to be validated “through a time-consuming, error-prone, and complex process involving several steps. The process generally encounters high defect rates, resulting in lower accuracy of contracts being loaded. This, in turn, results in backlogs and high turnaround time. The rework rate is also high due to multiple manual hand-offs and errors.”

Wipro says its team “designed four robots and sequentially bundled them to automate 75-80% of the process.” The benefits included a reduction of “nearly 30 minutes of post-processing documentation,” improved productivity, and cost savings

The automation of processing tasks has been underway for many years now, with many different technology approaches. Is RPA old wine in a new bottle? Perhaps what’s new is that RPA pulls together many capabilities – from forms validation to customer call management to predictive analytics -- into a single type of initiative, also lending low-cost software to bring it all together.

Deloitte’s Walsh says RPA can do seven types of tasks for insurers:

1) Gather, collate and validate information

2) Synthesize and analyze structured and unstructured data

3) Calculate and decide what to do

4) Communicate and assist users, clients and customers

5) Orchestrate and manage activities

6) Monitor, detect and report operational performance

7) Learn, anticipate and forecast behavior or outcomes

Stay tuned to this space. RPA is the new kid in town, but we expect to see stories of implementations and lessons learned emerge over the coming months.

For reprint and licensing requests for this article, click here.
Artificial intelligence Machine learning Cloud computing
MORE FROM DIGITAL INSURANCE