Insurance companies around the globe are faced with challenges of decreasing margins. Increased competition from banks and brokerages is not making life any easier for the insurance companies. Expenses have been more than the revenue growth for many of the financial institutions. Cutting down expenses is one of the mandates for all the companies in order to maintain their profitability. At the same time, the companies' need for computing resources is increasing day by day due to the ever demanding customer needs, better analytics leading to better risk management and correct decisions, and increasing regulatory requirements.
Against this backdrop, Grid computing seems to be the key for many of the challenges faced by insurance companies.
Grid computing concept is defined by Ian Foster as the controlled and coordinated resource sharing and problem solving in dynamic, multi-institutional virtual organizations. This sharing of resources, ranging from simple file transfers to complex and collaborative problem solving, is accomplished within controlled and well-defined conditions and policies. The dynamic grouping of individuals, multiple groups or organizations that defined the conditions and rules for sharing are called virtual organizations.
Grid computing focuses on the resource sharing, coordination and high performance orientation. It is all about resource sharing by integrating services across distributed, heterogeneous, dynamic virtual organizations formed from disparate sources within a single institution and/or external organization.
The concept of virtual organization is the key to Grid computing. A virtual organization is defined as a dynamic set of individuals and/or institutions defined around a set of resource sharing rules and conditions. All these virtual organizations share some commonality among them, including common concerns and requirements.
Why Grid Computing?
Grid computing provides consistent, inexpensive access to computational resources (desktops, servers, supercomputers, storage systems, data sources, instruments and people) regardless of their physical location or access point. As such, The Grid provides a single, unified resource for solving large-scale compute and data intensive computing applications. A look at some of the numbers makes the argument in favor of grid computing all the more compelling and convincing.
Over a 24-hour period, the average UNIX server is actually “serving” less than 10 percent of the time. For mainframes, the figure is about 40 percent; and for desktops, less than 5 percent. This means that an average desktop is idle for an unbelievable 95 percent of the time. This translates into hundreds of millions of dollars that could be saved in the form of valuable processing time by the use of Grid technology.
One of the compelling reasons for enterprises to adopt Grid computing is to achieve infrastructure and information virtualization. Today's industry, wrapped up with the idea of discrete applications running on discrete computers, is constantly haunted by issues of capacity planning, efficient management of system resources and robust performance of various systems (see insert 1).
The Department of Trade and Industry (DTI), UK has invested £1 million in the IECnet program in order to promote grid computing and raise the awareness of the benefits of grid computing and overcome board level ignorance. Grid computing would facilitate the delivery mechanisms for existing services.
Some of the benefits of Grid computing include:
- Aggregating the performance of many systems to build a virtual super computer.
- Creation of a virtual organization enabling sharing of the applications and data — linking together of information located at multiple locations.
- Application optimization.
- Optimizing and managing the workload across multiple resources.
- Effective and efficient processing — stignificant reduction in processing time.
- Reduce the total cost of ownership.
The financial services industry can greatly benefit from the use of Grid computing. A few insurance companies (See insert 2 ) and employee benefit providers (see insert 3) have already introduced Grid computing and greatly benefited from the technology. The Royal Bank of Scotland (RBC) has replaced an existing application and Grid-enabled it based on IBM xSeries and middleware from IBM Business Partner Platform Computing. The integrated solution allowed RBC Insurance to reduce by 75 percent the time spent on job scheduling, and by 97 percent the time spent processing an actuarial application. Based on that success, RBC Insurance is now looking to expand the IBM and Platform Computing solution across additional applications and business units to improve efficiencies and to increase customer satisfaction.
“IBM and Platform Computing Grid-enabled our valuation application and supporting infrastructure for immediate results,” said Keith Medley, head of insurance technology at RBC Insurance. “With the integrated solution, we have been able to reduce a 2.5 hour job to 10 minutes, and an 18 hour job to 32 minutes. We are now looking to move to a production environment. By virtualizing applications and infrastructure, we anticipate being able to deliver higher quality services to our clients faster than ever before, which will significantly impact our competitive edge.
Hartford life is among the first life insurance companies to implement Grid computing. The complexities of the variable annuity products called for efficient risk management, which translated into a demand for high computing power. Management realized that Grid computing could solve their pressing business needs. Using Condor, a Grid computing software from the University of Wisconsin, Hartford completed the project on its own. Now they are able to simulate market behavior, policy holder behavior for the next 15-20 years and map risks. Grid computing has helped Hartford life in many ways. It has helped them to build a very sophisticated risk management capability. Other benefits include scalability with increasing volume, stability while expanding the network and cost savings. The challenge it faced was lack of expertise on the subject matter since it is a newer technology. Though Grid computing is a viable technology, organizations need to possess right engineering acumen for a smoother implementation.
Global HR and benefit plan provider Hewitt demonstrated that grids can solve their business problem. Hewitt had an application hosted on a main-frame system that performed complex pension calculations for 5.5 million of Hewitt's customers. Usage volume would fluctuate wildly due to rumors on mergers, acquisitions or early retirement programs circulated, and volume could double without warning. These pension calculations at such high volumes were consuming a lot of computing power of the host system and becoming very expensive to run on a main-frame. The application made an ideal grid candidate for another reason, too. Apart from querying the database in an efficient way to collect the data needed to perform its calculation, the application interacted very little with other IT systems and data hence the problem could be divided into smaller pieces and can be executed at different locations. In Hewitt now the grid works like a giant CPU and returns the end result of the pension calculation query to the mainframe. The introduction of grid computing resulted in a reduction of 90% of the transaction cost.
Most of the Grid computing solutions are home grown. Vendor support for Grid computing has not matured yet. Grid computing has limited areas of deployment. Not all problems can be solved using Grid infrastructure. Grid is more suited to situations where the problem can be split into smaller problems. Firms have also found implementing Grid computing has had its problems and technical hurdles have been high.
An ideal grid should ensure that the resources of the different systems are used seamlessly and without affecting the performance of the individual users systems. It also goes without saying that enough security features are built into the system. Grid technology planned and executed properly could bring in enormous benefits to the insurance industry. Optimal usage of the existing resources will be the driving force that will ensure competitive advantage and provide leadership position.
About the Authors
N.R. Veeraraghavan is a senior consultant in the insurance domain of Infosys Technologies Ltd, India. Nandha Kumar is a consultant in the insurance domain of Infosys Technologies Ltd, India.