Airplanes, U-Boats & Data Signals: Data Analysis Lessons For Operational Improvement
At this month’s Technology in Action Conference, our Director of ProAct engagements, Jeff Zazzara, and I will be giving a presentation that is at least partially inspired by an article we found on LinkedIn, “The Story of the Way Big Data Shaped World War II” by Michael Moritz, the Chairman of Sequoia Capital. Not only does the article combine two of my favorite topics, history and data, but it also reveals an important data analysis lesson: the importance of data signals.
In the world of data analysis, rows in a spreadsheet are termed “cases” and columns are “signals.” Adding cases to a dataset, such as recording kWh consumption over time, reduces the likelihood of outliers and random variance. However, the most powerful way to find valuable insights from a dataset is to add signals. Adding signals creates context for the data; it can reveal new relationships (correlations) and enable key inferences which drive action.
In the military context of The Battle of the Atlantic in WWII, the Allies’ dataset had many cases of attempts to sink German U-boats with air-to-sea depth charges, but each case had only one resulting signal: whether the U-boat was sunk, damaged, or unharmed. Identifying specific actions to improve the success rate of these attacks was very difficult without additional supporting data signals. By adding a variety of signals to their dataset – trigger depth of the depth charge, time elapsed from enemy sighting to engagement, and even color of the aircraft – British Intelligence was able to build context into their data and identify potential tactical improvements. As a result, they painted black airplanes white, changed the trigger depth to 25 feet, and improved the success-rate of attacks on submerged U-boats from one percent to seven percent.
In the HVAC/R context today, I see maintenance managers and energy specialists realizing the value of data signals and taking steps to harness their power. One successful Emerson partner improved their preventative HVAC maintenance scheduling by using the E2 management system to create signals from the number of high space temperature alarms and the percentage of units at a store not meeting temperature differential targets. Armed with new data, they targeted stores and units most likely to fail during the peak summer months.
Emerson’s technologies and solutions, such as CoreSense™ technology and the E2 management system, are creating, and will continue to create, new data signals and new opportunities for operational improvements. The key to creating savings through Smart Data is maximizing the number of data signals available to you and then identifying which ones best drive specific key performance indicators.
Emerson Climate Technologies customers who will attend TAC in a couple weeks will hear Jeff and I further explore the use of data analysis in WWII (ask to hear the bullet-hole story), how maintenance teams are using data signals today, and how Emerson can help you develop a signal-rich improvement plan.
But, I am also interested in hearing from experts outside of the retail market. How are you using signals to find meaning in data and improve your business? What are some of the difficulties you have encountered in your endeavors?
Emerson Climate Technologies, Retail Solutions