Robots are taking over the world. They are no longer sci-fi fantasy and today robots powered by artificial intelligence (AI) can clean your house, educate the young, care for the old, and fetch your wine and slippers. Reports suggest that the global robotics market will be worth £100 billion+ by 2020.¹ This chapter concentrates on the specifics of robotics, rather than the full spectrum of artificial intelligence (AI), which incorporates neural networks, data science and machine learning, and the many other categories in between, which constitute the world of AI.

Robotics is a powerful, tactical tool that, utilized correctly within a clear architecture and roadmap, can deliver significant, immediate cost and quality benefits, particularly against repetitive, manual tasks and legacy connectivity. However, robotics is not a strategic, enterprise-level transformation tool nor a silver bullet.

Today the insurance industry is just starting out on the robotics journey and very little is understood about what robotics can do for insurance and how. Even less is known about what is commercially viable today and what is still a pipe dream.

This chapter discusses the different forms of robotics available to insurers and when to use them, what is commercially viable, and why some projects fail. It will also consider the importance of people and robots working together, with governance and oversight over the robots. Six key areas are examined.

When to Deploy Robotics in an Insurance Organization

The main characteristic of a robotics-ready environment is where the concept of quality prevails over quantity, volume, and cheap labour. In a successful robotics environment, productivity increases by relegating non-value-added (NVA) work to robots and concentrating human activities on intelligent work in parts of the business process that are customer-centric.

Robotics can have significant impact when people are obviously doing things by rote in order to fill in the gaps or when business is moving too quickly for standard solutions. It is particularly effective in situations where:

• legacy systems and processes (particularly multiple legacies) inhibit productivity;
• APIs or services don’t exist;
• service/API integration is too slow or too expensive to deliver ROI;
• problems are currently solved through training, documentation, and process outsourcing.

In all these situations, it is external customer demand that has overtaken traditional interactions and processes from where and how they originated, often prompted by more agile competitors and InsurTech startups redefining industry practices. The traditional insurer is left behind and has to race to catch up, otherwise others will step in and their customers will disappear.

However, robotics does not significantly re-engineer business processes enough to deliver the full benefit of transformation and these tactical quick wins should be recognized as valuable only within a long-term strategic transformational plan.

What are the Typical Benefits Robotics Can Deliver?

The benefits that can be derived from using robotics within an insurance environment centre on operational improvement, productivity increases through efficiency and effectiveness, increased quality of execution, performance recording, and improvement in customer and staff experience.

• The first benefit insurers gain from robotics is an improved operational resilience through an ability to manage resource capacity during operational peaks and troughs.
• Typically a robot produces three times the output of one full-time employee and works 24/7 without benefits at 50% of the cost of an off-shore human resource. As a result, average time to complete a process can be reduced by 40–60%, both in terms of elapsed time and man-hours, thus reducing costs by 30–80%.
• Instilling 100% accuracy and consistency eradicates the costs associated with errors and omissions and rework.
• The creation of a full audit trail eases regulatory compliance and facilitates risk management.
• Robotics enhances customers’ experience by reducing wait times and accuracy of execution.
• Simultaneously, it raises staff satisfaction by reducing monotonous working practices.
• By using robotics, insurers can accelerate time to value and return on investment. A reasonable expectation is for a robotics process to be designed, configured, tested, and deployed in around six to eight weeks and ROI to be realized in three to six months: not the timescales associated with traditional project business cases.

The Different Types of Robotics and When and How to Use Them

The world of robotics is rapidly evolving: in the products available, the terminology used to describe them, and the burgeoning list of vendors in the market. No one product is the answer to all situations, and insurers should ensure they carefully consider the applicability of any one tool or type of technology when establishing a business case for a project. It is also crucial to establish how it integrates with other technologies such as business process management, case management, and AI and exploit what they bring, rather than trying to recreate them. In many cases it is the integration with these other technologies (and the ongoing maintenance) in a larger transformation that can determine the strength of the robotics business case. A unified architecture is a major advantage.

Virtually all recognized robotics vendors maintain that the majority of their revenues come from customers in the US$1 billion+ or even US$10 billion+ turnover category; the economics of robotics is very much about scale and its geography is mainly large international insurers.

However, while it is certainly true to say that the established market for robotics is larger insurers, vendors are almost always either small (organizations with <US$20 million turnover) or smaller autonomous divisions of larger technology vendors. Careful due diligence is a watchword for larger corporations putting strategic reliance on the capabilities, R&D, future direction, and commercial stability of considerably smaller vendors, some with exposed commercial futures.

As the established analyst companies focus on the robotics market, some conformity of terminology is emerging to facilitate better comparisons and the opportunity to explain the application of different tools to different business problems.

Identifying which Processes are Viable for Robotics

As described previously, the correct technology has to address the correct business problem. To maximize success, there are some key minimum characteristics to look out for when selecting the business process to be addressed:

• It should be predominantly rule based and not depend on human judgement.
• It should be supported by digital, preferably structured data.
• It should be functioning and stable.
• It should save significant time and complexity.
• It should be reusable and scalable across the business, including multiple geographies, with little or no adaption.
• For a proof of concept project, it is key that the process leverages the key systems of the company.

Examples of Processes that Can be Automated with Robots

Process selection typically falls into two categories: generic activities that can usually be undertaken in all organizations and those specific to the insurer. The former category includes processes such as users logging onto systems and applications, data extraction activity (e.g. from forms or websites) and the creation of an audit trail for operational and regulatory compliance purposes.

Processes specific to insurance might typically include:

• Address or payment changes and fund transfers
• Policy amendments
• Compliance checks
• Background validation, e.g. credit checks, fraud detection
• Data capture and triage for underwriting, particularly with extensive schedules originating in spreadsheets
• Autogeneration of documents
• Payment reconciliation
• Month end consolidation.

The Human Factor and the Role it Plays in Success

There is no doubt that robotics, alongside AI, elicits the same concerns over livelihood protection as with Luddites of the Industrial Revolution. Although replacing humans with automation technology has continued unabated since the nineteenth century, robotics, by its very name, raises particular human concerns and a successful program has to address these issues very thoughtfully.

Insurers must carefully explain the strategy of “taking the robot out of the humans”, which will focus their staff on more complex, intellectually challenging activity and ensure humans and robots work in harmony together. Redirecting staff to activities with a higher requirement for judgement and an emphasis on interacting with customers rather than repetitive, rote activity, not only drives up NPS² scores with policy-holders but empowers staff and improves morale.

Success is achieved through a bi-directional approach to achieving buy-in across the business. The biggest failure for any project is lack of committed top-down sponsorship, so this has to be very visible to all concerned. This, together with strong governance and the deployment of reusable standards across the business, is an essential component to achieving objectives.

However, this has to be combined with bottom-up behavioural change programs. These powerfully emphasize the personal benefits accruing to staff to prevent a negative attitude to adoption. Alongside expert team coaching on new skills and the celebration of success, they raise the probability of success at the grass roots level.

Summary

Robotics is not the answer to every problem and cannot deliver enterprise transformation. It is a rapidly evolving technology with enormous potential to shape businesses quickly. Care must be taken in choosing vendor partners, the right tools, and the correct targets for robotic deployment. However, with sensible strategy, informed execution, and a strong communication plan, robotics can add significant business improvement, future agility, and overall value to the operations of every traditional insurance organization operating today.