The Rise of Machine Learning

Big-Data-4-compressorSchneider Electric is applying machine learning to its growing Internet of Things platform to predict when its products—which range from sensors to programmable logic controllers—may fail, so that they can be serviced and avoid costly downtime.

The French company, which has U.S. headquarters in Andover, Massachusetts, also applies machine learning to its building-management systems to make structures more eco-friendly and comfortable for occupants.

Atlanta-based GeoDigital is using machine learning to apply its detailed three-dimensional models of the world to a number of industries. For utilities, GeoDigital can precisely locate assets as mundane as power poles and then use machine learning to recommend maintenance steps and schedules. Geo-Digital’s 3D-mapping programs also are feeding the revolution in the automobile business known as self-driving.

For vehicles to be truly driverless, yet safely complete their routes, they’ll rely on a cloud of data and a flurry of instantaneous communications that will keep them apprised of all the conditions around and within the vehicle—both static, like the contours of roads, and dynamic, such as traffic conditions and whether a pedestrian is darting in front of the car at that moment—and advise them how to respond.

“What used to be acceptable becomes an impediment if you need a vehicle to make decisions on its own about whether to make a maneuver immediately, or prepare for traffic conditions a few miles away,” says Anupam Malhotra, senior manager of the connected vehicle for Audi of America. “Machine learning is how we close the gap to make truly automated driving possible.”

Big Moves in Big Data
Machine learning is just one way to leverage Big Data. Here are examples of how CEOs are using a variety of tools with Big Data to advance their companies:

• At TransUnion, the Chicago-based credit-scoring giant, a new data-powered product called Credit Vision is providing lenders with more granular and nuanced assessments of a consumer’s risk factors, which is helping them unearth more high-quality prospects and make more sound loans.

“This also is helping us increase our footprint with the lenders we serve and boosting the ‘stickiness’ of our other products, such as fraud solutions,” explains TransUnion CEO Jim Peck. “That helps us maintain and grow our positions with our customers.”

• Edmunds.com, the automotive-information web site, has been rolling out a new product called Car Code, which allows consumers to directly text dealers about vehicle pricing, availability, features and questions. Predictive analysis of the resulting millions of conversations turns up key findings that will help make the next texting exchanges more effective.

Big-Data-5-compressor“Some of our early analysis showed that if a text conversation included the words ‘wife’ or ‘spouse,’” reports Avi Steinlauf, CEO of the Santa Monica, California-based company, “the final outcome was four times more likely to be an actual sale. Big Data provides a lot of actionable information like that.”

• Neil Clark Warren returned to his previous CEO position and boosted the importance of data massaging to help another Santa Monica-based company, the dating site eHarmony, recover against key competitors such as Match.com. Actionable insights from data analytics included the fact that dating candidates from outside the U.S. are more comfortable being matched with someone who smokes cigarettes or drinks booze than eHarmony’s American members are, so it broadened the matching algorithms for overseas users.