Prepping Your Crystal Ball: How Predictive Analytics and Smart Machines Can Benefit Business
Companies can use predictive analytics to support executive decision-making, increase productivity, and drive economic benefit.
Knowing the future would make managing a business much easier, and predictive analytics is a growing branch of computer technology that tries to accomplish just that. Analytics systems use "smart machines" equipped with specialty software, data-collection hardware, powerful servers, data-mining tools, text analysis, and complex mathematics to forecast the probability of future events by examining historical and current patterns in big data.
Smart machines equipped with predictive analytics tools can show likely future outcomes and possible courses of action for business leaders.
Analytics tools are easy to use, even for non-IT employees, and can aid productivity by determining ahead of time when machines’ parts will need maintenance or replacement.
GE is incorporating predictive analytics into many of its products
Predictive analysis still requires intelligent human judgment to interpret the results and take decisive action.
In some cases, the future is now, as companies are already beginning to amass analytics data. However, the C-suite should be forewarned that handling such a massive volume of information may require increased storage and processing capacity. Moreover, businesses may need to call on data scientists or their brethren to make the most of predictive models.
As GE CEO Jeffrey Immelt said, combining computing power with big data is "beautiful, desirable, [and] investable" because it could "drive massive economic benefit."
According to Harvard Business Review, "Predictivity will eventually connect all GE's machines to the cloud (no small feat, given that some business units, such as health care, have thousands of products, each with its own complex software needs and legacy systems), enabling them to talk to one another, learn from historical data, and provide predictive information to help eliminate unplanned downtime and otherwise improve efficiency."
This offers businesses multiple advantages, according to Gartner: Analysis happens more quickly ("in hours or days"); it's tailored to a company's needs by emphasizing business-relevant findings; and analytics tools are easy to use, even for people who aren't tech-fluent. It also helps customers by increasing reliability and productivity.
Rich Carpenter, CTO of GE Intelligent Platforms, described predictive analytics for The Manufacturing Connection in a step-by-step fashion: "collect data –> store data –> analyze –> diagnose why things are not 100 % –> [make] recommendations." He continued, "Now you can start planning downtime — not wasting time and dollars due to unstable operations."
GE's new Drilling iBox is one example of a tool that provides predictive analysis. Equipped aboard offshore oil platforms, it tracks the status of complex equipment by using sensors to "report to operators topside on critical information — valve positions, temperature ranges, stresses, and conditions." Operators can then use predictive analytics based on the data received to determine "exactly when parts will need maintenance or replacement," thereby reducing (potentially hugely expensive) pauses in productivity through pre-emptive upkeep.
Beyond large machinery, GE found in its Industrial Internet Insights Report that health care also stands to gain "improved clinical, financial, and operational outcomes" with the use of predictive technology, "including improved diagnostic speed and confidence (named by fifty-four percent of respondents); reduction in patient wait times and length of stay (fifty-six percent); and better clinical outcomes and patient satisfaction scores (fifty-nine percent)." Eighty-four percent of those surveyed also agreed that health care providers who embrace a predictive analytics strategy in their approach to patient care "will outpace their peers in the marketplace."
Your business can get valuable information from predictive analytics, but this initiative will require hardware preparation. GE has embedded sensors into 250,000 "intelligent machines," as InfoWorld has reported, and analyzed 5,000 data points per second on its GEnx jet engine to optimize flight times. According to GE Global Research CTO Mark Little, that initiative includes monitoring a fleet of 25,000 engines to help predict failures before they happen.
For other companies, data could just as easily come from factory controllers, sensors in company car fleets, or even customer devices (while taking privacy into account). For example, office imaging machines could automatically transmit information, including usage rates and problem codes, to secure off-site locations, enabling dealers to schedule a single service call once they’re armed with the right information, parts, and supplies.
To handle the data volume, you may need increased storage and processing capacity.
Predictive analytics mainly uses historical information, which can lead to limitations. Many strategic decisions concerning new markets, product types, or practices may not profit as much from this retrospective analysis. Also, good predictive work relies on mathematical models, and while there are preconfigured ones available, more complex issues and questions may require experts to design custom models.
In addition, hiring skilled staff may be necessary. Although the tools themselves are easy to use, interpreting the data requires expertise. "Now there is much more data available to the clinician," said Dr. Christopher C. Colenda of the West Virginia United Health System in the Industrial Internet Insights Report. It's important to have "the right people with the right skills to interpret the data — biostatistics, epidemiology, health informaticists, other health professional clinicians, and so forth. Caring for patients is now a team activity, and learning to work in teams is an important skill for physicians to acquire."
This just serves as a reminder that analytics will not replace human management, but aid it. Management legend Peter F. Drucker called computers "morons" in McKinsey Quarterly back in 1967, writing, "The stupider the tool, the brighter the master has to be — and this is the dumbest tool we have ever had." Computers may have become much smarter since, but the need for intelligent leadership is as strong as ever.
Erik Sherman is a journalist and author whose work has appeared in such publications as The Wall Street Journal, The New York Times Magazine, Newsweek, the Financial Times, Chief Executive, Inc., and Fortune. He also blogs for CBS MoneyWatch. Sherman has extensive experience in corporate communications consulting and is the author or co-author of 10 books. Follow him on Twitter and circle him on Google+.