A U.S. Marine walks through the door of his house. He drops his duffel, and picks up his young daughter, wraps his arms around her and twirls her, laughing. He’d been in Afghanistan for almost a year, and now he has two years to spend at home with his family. How did he get home safe?

This brave dad is just one of more than 200,000 Marine Corps warriors who must be ready to deploy at a moment’s notice, heading straight into danger, from conflicts in Iraq and Syria to disaster relief in Puerto Rico and Texas.   

The challenge

Real Life Game of Risk

Marines are led by commanders who juggle many responsibilities and requirements. Some missions are enduring (e.g., multi-national exercises with allies; Marine Expeditionary Unit deployments), but flare-ups happen worldwide. These leaders and their staffs must use their experience, skills, and a massive amount of data to schedule, plan, and send troops all over the globe. And even with all the high-tech weaponry and analytics tools in the world at their disposal, much of the planning and scheduling is done with spreadsheets and sketched out ad hoc on whiteboards. Bureaucracy gets in the way of strategy. They’re playing a real-life game of Risk, with the highest stakes imaginable: making sure Daddy gets to walk through that door.

Transformation story

Predictive Analytics for People

The Marines are now using the same Watson predictive analytical capability that diagnoses the health and readiness of military vehicles and applying natural language processing and data science to the organization and deployment of people. The force management tool includes SPSS--a statistical package that can perform complex data analysis with simple instruction--for planning; and Watson Explorer for natural language processing that can read and analyze commanders’ notes.


The pilot, running on the Secret Internet Protocol Router Network (SIPRNet), applies parameters including depth-to-dwell ratio (the amount of time a Marine spends at home versus time he/she spends deployed); the types of units (infantry, aviation, etc.) and their respective readiness, and other variables, and analyzes all these factors. When a hotspot flares up or disaster strikes, planners can use this newly developed readiness tool to pick the right units to deploy for each situation while balancing the need to allocate forces for future missions and while minimizing impacts to future readiness. It means determining which units are needed to keep driving ISIS out of the Middle East, and which to send to rescue survivors after an earthquake in Japan.

Results story

Augmented intelligence

Because the system uses augmented intelligence, each one of those moves can be toggled to show the ripple effects through five years of planning—the standard planning window. It’s all done with a click and a drag. In the past, it was pages and pages of Excel spreadsheets and human analysis alone. The custom dashboard allows commanders to see the readiness level (skills, length of current deployment, distance from battle zone) of the units on a single pane of glass, to aid their decision-making, getting the right people to the right place at the right time.

In the short term, the Marine Corps and IBM hope to expand the pilot to cover units worldwide. The goal is to implement an all-force planning tool that gives decision-makers the data to understand where to invest in additional training and recruitment, stay prepared for emergencies and unexpected conflicts, and keep troops healthy, ready, and safe.


The real breakthrough will come with optimization and prediction: technologies like this will eventually go beyond analysis and be able to actually recommend courses of action, which means faster and more precise data-driven decision making. The system could tell a commander and staff the best moves to make before they know they need to make them, even predicting the need for a surge in recruitment of new Marines who won’t even be ready to deploy for years. In the end, though, it’s all about getting even more dads—and moms—home safely.

Learn more about using AI for readiness - including how to make the business case.