Analytics Magazine Analytics Magazine, January/February 2014 | Page 10

EXE CU TIVE E D G E Now we’re at the beginning of a long rally race of analytic improvements – from newer, better data to smarter, faster algorithms. Enhancements will include broader, more scalable platforms and will access unique sensors – spectra, spatial, temporal – and micro/macro levels of structured and unstructured data. 10 | unique sensors – spectra, spatial, temporal – and micro/macro levels of structured and unstructured data. All that will help generate insights into individualized, massively personalized and localized information while getting even more power out of grouped predictive parameters. For example, if you have operated a car or other moving object, you undoubtedly have been assessed for your risk of loss as an operator of that vehicle and, in some manner, likely have been insured. In the past, that insurance-based risk assessment has blended wide bands of information on a few historically available generic characteristics to achieve a generalpurpose estimate of prospective loss risk estimates. In the future, that historical benchmark will be segmented into ever-more granular and accurate assessments. Those will then again be reinvented, recombined and refined to enhance the data-driven process, culminating in an adaptive analytic that adjusts expectations to the level of risk in each operating scenario encountered or intuited. In the world of big data and bigger analytics, insurers will come to view vehicles as instrumented platforms and operators as real-time learners whose risk may change over time. Operators may drive safer and make smarter decisions about moving between locations, or they may permit distractions into their cockpit (such as texting, talking, smoking, eating and so on). How you drive, when you drive, where you drive and how much you drive are all becoming part of the context in your individual risk profile. Some businesses apply detailed telemetry, routing algorithms, real-time weather and traffic alerts and driver/crew pairing models to manage more A N A LY T I C S - M A G A Z I N E . O R G W W W. I N F O R M S . O R G