How Predictive Analytics Is Being Used

Predictive analytics methods are being used for a wide variety of purposes. Below are seven key examples.

  • For making supply chain operating networks work better. Gary Neights is Senior Director at Elemica, which provides network solutions for global process industries. He notes that predictive analytics (PA) is being used to predict both the oversupply and the undersupply of finished goods as they flow through the global supply chain. “In one example,” he told Predictive Analytics World, “the predictive system drove inventory replenishment accuracy from less than 55% accuracy to greater than 80%….This drove a working capital savings of greater than $400K / year for one product.”
  • To increase customer focus and satisfaction. PA can help companies map “customer journeys” so firms can gain insight into customer preferences. That information can be used to predict which actions will lead to higher customer engagement.
  • To look at patient medical histories with an eye toward predicting future events and conditions, according to Tom Check, president and CEO of Healthix, a health information exchange vendor. The goal is to take action based on clinical data and apply algorithms to predict potential future risk.womens eye-series-1438496-639x852 (1)_edited-1
  • To help doctors plan and carry out cataract surgery. IBM and Bausch + Lomb are partnering to create an iPhone app that will “electronically manage patient information across an iPhone or iPad while hosting health-related data on IBM Cloud.” The system can relay recommendations in regard to a lens that can be implanted in the eye to treat cataracts. Surgeons may use it to “enhance surgical planning and improve patient outcomes.”
  • To improve the employee selection and management process. PA can be used to make the interviewing process more effective, to aid employee retention, and to locate the best external pools of talent.
  • To predict the failure of electrical and mechanical components and, therefore, system failures. The idea is to “reduce unscheduled maintenance and increase operational availability of critical assets,” according to Jeffrey Banks, Department Head at Complex Systems Engineering & Monitoring at The Applied Research Laboratory at Pennsylvania State University.
  • To predict errors or other failures on the production line before they happen. Jabil, a design and manufacturing solution provider, is using Microsoft’s Azure services to “analyze millions of data points from machines running dozens of steps throughout the manufacturing process.”

Leave a Reply