A Day in the (Future) Life of Predictive Analytics
A look into a normal day in the near future, where predictive analytics is everywhere, incorporated in everything from household appliances to wearable computing devices.
By Lars Rinnan, CEO, NextBridge
One day in the near future.
While eating breakfast, a message appears on the screen of your refrigerator that it is 87 percent likely that you will run out of milk today given today's date and your activity log. You agree to add milk to the shopping list shared with your preferred supplier of household items. The items on that list are delivered to your door an hour after the first family member is registered by the refrigerator as present in your house.
You go out to your car, which, thanks to biometric scanning, recognizes you and opens the door and lets you press the start button. You say you're going to the office and the car responds that "You are 18 minutes later than usual, so it is highly likely that the normal route is not the fastest, and that alternate route 2 will save you 12 minutes." The recommendation is based on the Public Transport Agency's analysis of log traffic data of all roads in the city. You accept alternate route 2 and it comes up in your head-up display.
During the drive, your music streaming service plays music it predicts you'll like, based on past music you've played in your car. While driving, the vehicle's sensors sends traffic data to your insurance company, which uses dynamic pricing to calculate your insurance premium. Because you are an accountable driver, this has reduced your premium by several hundred dollars over the past six months.
The vehicle information system notifies you that the battery charger is likely to fail in the coming weeks and suggests you make an appointment at the garage. The system provides three alternative dates, which are coordinated with your preferred garage booking service and your Outlook calendar.
Your car dealership recommends you replace your car during the next six months for optimal economic car ownership based on your traffic data and predicted supply and demand for similar car models. Your bank has tailored an offer for a car loan based on your current and predicted private financial situation and payment history. Other car dealers and car-related services have also predicted that you are in the buying mode for the new car, and because this is an important decision, you have opened up to receive such offers for a specific period.
As you drive past a car dealership, an offer on one of its new models is displayed on your car's communication system. The offer is so interesting that you decide to stop and take a look at the model. The car has the latest generation of battery technology with a range of 1000 km and accelerates from 0 to 100 km/h in less than 3,5 seconds.
During the test drive, the extreme acceleration makes your heart beat so fast that your personal health data sensor triggers an alarm. The health data sensor is integrated into the strap of your wrist watch. This data is transferred to your health insurance company, so you say a prayer that their data scientists are clever enough to exclude these abnormal values from your otherwise impressive health data. Based on such data, your health insurance company's consulting unit regularly gives you advice about diet, exercise, and sleep. You have followed their advice in the past, and your performance has increased, which automatically reduced your insurance premiums. Win-win, you think to yourself, as you park the car, and decide to buy it.
Most of this short story is already commercially available or technologically possible. The global volume of data continues to grow exponentially, and the ability to draw knowledge out of these vast amounts of data is also becoming widely accessible.
Prediction based on large amounts of data will change how business and government operate. Sensor data about your power consumption triggers differentiated pricing depending on the time of day consumption occurs. This balances the power consumption throughout the network, avoiding unnecessary development, resulting in both financial and environmental benefits.
Your personal preferences and activities shared on social media enables more personalized and relevant marketing. (Perhaps this will lead to the long overdue death of mass marketing.)
Improved prediction of demand for public services (such as roads, schools, and sports arenas) will mean that these are in place when the need arises, not five years afterwards.
Prediction of tax fraud and insurance fraud will reduce unwarranted government payments.
The applications for predictive analytics of large data are virtually unlimited, and this story is just an appetizer. The yield will be considerable. The question is, will there be enough talented data scientists to handle the data and will management be willing to explore new possibilities?
Lars Rinnan is CEO at NextBridge and founder of the Gurus of BI conference. You can contact the author at