Organizations in a variety of industry sectors are showing keen interest in intelligent software tools that automate basic business processes and support functions.
Smart "robots" are able to execute routine and mundane tasks faster, cheaper and more accurately than humans.
As the technology gains traction, adopters are discovering that the benefits of automation extend far beyond cost savings. Enterprises are increasingly seeking to use different types of intelligent computing to address critical industry challenges related to areas such as regulatory compliance, innovation and data man agement.
For insurers, smart computers have the potential to streamline a wide range of back-office processes around claims management, invoicing and customer service. In addition, combining automation tools with software that mimics human reasoning may allow insurers to tackle the thorny and longstanding problem of moving backlogs of legacy data from disparate forms to digital platforms.
While the term "artificial intelligence" is often applied broadly to any machine that performs sophisticated task requiring some level of knowledge or expertise, smart computers comprise a range of capabilities with clearly defined areas of focus and competence.
'Robot' software
At one end of the spectrum, robotic process automation (RPA) software is probably the most mature area of smart computing. Operating as a "virtualized full-time employee" or a "robot," RPA tools manipulate existing application software in the same way that a person follows rules or criteria to process a transaction or complete a routine task.
RPA tools don't replace existing client or service provider applications, but operate with those applications to perform tasks based on clearly defined rules and criteria. As such, they must be "taught" their jobs and have no "learning" capability in terms of drawing inferences based on past experience.
The most apparent benefit of RPA is dramatic reduction in operating costs. A single robot can perform the routine administrative tasks of multiple human workers. That said, rather than replacing, say, 40% of the people in a 100-person department, RPA software takes on 40% of the work done by every individual on the 100-person team. This creates an opportunity redeploy staff to take on more value-adding roles. Alternatively, labor costs can be reduced through attrition and by using RPA potential to increase productivity.
Insurance companies are using robotic process automation software to improve operational efficiency around claims, membership and customer service, as well as in generic business areas such as finance and human resources. Beyond delivering immediate cost savings, the technology can drive significant and ongoing business improvement. Once implemented, RPA produces metrics and data that provide critical insight into operational performance. A traditional Six Sigma program would involve months of data capture, analysis and reporting to achieve what RPA does essentially from day one as a matter of course. With a robot, in other words, the measurement and analysis of performance are built-in.
As such, the data collected by robots working in an enterprise can help optimize the performance of those robots. Indeed, one of the key challenges surrounding RPA is optimizing the robots' use rates — while they can work 24/7/365 without interruption, interdepartmental backlogs can result in robots standing idle. To address this challenge, data from RPA software can be applied to ensure enterprise-wide integration, balancing of workloads across the enterprise and agile responses to peaks and valleys in demand for robotic resources.
Continue reading …
IBM's Watson is an example of a computer with cognitive capabilities. (Photo: Wikipedia)
'Cognitive' software
Further along the spectrum of smart computing, "cognitive" applications apply natural language processing and pattern recognition capabilities to analyze data, draw conclusions and learn from experience.
Much like a human can scan a newspaper article for a few key words and identify whether it's from the sports or business page, a cognitive computing application can detect patterns and reach a conclusion.
Cognitive tools must be fed data to build a base of knowledge, and require specialized, scenario-based training to develop a capacity for logical reasoning within a defined set of parameters. An example of cognitive capabilities is IBM's Watson, which uses natural language processing, hypothesis generation and evaluation and dynamic learning to analyze unstructured data.
Combining the two
Perhaps the most exciting opportunity around smart tools for the insurance industry lies in the linkage of existing robotic process automation capabilities to cognitive "thinking" applications.
Combining these tools has the potential to address the onerous problem of taking massive backlogs of legacy data stored in claims, account updates and invoices, and moving that data to digital platforms.
On its own, robotic process automation can only tackle part of that task. While they perform specific tasks rapidly and accurately, RPA tools require highly structured input and explicit training. When encountering an exception to the rule it's been taught, an RPA tool is stymied. Legacy data, meanwhile, is unstructured and full of exceptions — a policy number might be in one box in document and in a different box in another. As a result, the processing the data in legacy paperwork doesn't lend itself to specific, consistent and repeatable rules.
Cognitive applications, meanwhile, do have the capability to manage exceptions, discern patterns and identify irregularities — the capability needed to extract insurance policy numbers from different boxes on different forms. This creates an opportunity to deploy cognitive solutions on the back end to analyze unstructured data contained in disparate forms, invoices and policy documents stored in legacy systems and extract the relevant bits. RPA solutions on the front end process that data 24/7 and transfer information to digital platforms.
While development of the synergy between robotic process automation and cognitive is still in early stages, many insurers are exploring the possibilities. Organizations that invest wisely in institutionalizing their existing robotic process automation platforms will be positioned to achieve maximum benefits by implementing cognitive tools as just another training task for a robot.
Paul Donaldson is a director with Addison, Texas-based Alsbridge, a sourcing advisory and consulting firm.
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.