How Prescriptive Analytics Works in the Real World
In our first blog post on prescriptive analytics, we described what it is and how it works. In this second post, we’re going to explore a few practical applications of it.
Crew recovery operations at one of the world’s largest airlines
Continental Airlines (now United) faced a challenge many large airlines have, the complex scheduling of crew members. Flight delays and cancellations at one airport cause delays and cancellations at others. Since airline schedules require routing of aircraft and crew on continuous paths, crew could find themselves unable to board their next scheduled flights and leave planes without the correct levels of staffing. The negative effects on airline revenues and customer satisfaction are significant.
Continental developed a decision-support solution aimed at returning its crew to their scheduled assignments as quickly as possible. Constraints including government regulations, employment terms, and pilot qualifications to fly specific aircraft placed critical sets of limits on crew recovery operations. They even needed to consider crew members’ preferences for assignments for the sake of morale and contractual obligations.
All of these constraints as well as decisions such as canceling or delaying flights provided input to the crew recovery optimisation model. The target in crew recovery was to minimise the added costs of covering the remaining flights in a disrupted schedule (in addition to the cost of implementing regular schedules). The choices in the crew recovery problem were the specific assignments of crew members, such as having a crew member work a flight or be “deadhead”, to travel as a passenger to arrive at an airport for their next assignment.
In the first year alone, Continental estimates it saved USD 40 million in four major disruptions the crew recovery solution. And in the immediate aftermath of the tragic 9/11 attacks, the new system proved critical, allowing them to produce new schedules with the crew recovery solution. They were able to return to normal operations before any other airline.
McDonald’s manpower planning solution
McDonald’s APMEA (APAC, Middle East, Africa) business needed to standardise the quality of their manpower planning process across all their outlets. With Decision Optimization, they developed a planning solution able to manage on multiple levels, from individual outlet to the national or even regional level.
The interactive and collaborative platform they developed reduced planning time from one day to less than one hour. Their staff utilisation increased by 8%. They remained 100% compliant with their legal and company regulations. And they were able to do what-if analyses for new stores and constraints effortlessly.
A massive retailer optimises shelf space
A large retailer with thousands of stores in many different formats with different needs and rules based on their location previously had rules for specific stores, determining floor plans on an ad-hoc basis.
An optimization solution they developed would take the rules that planners would input into a central rule management system, match it with store performance data, analyse this data and provide planners with detailed planograms showing shelf plans by every store’s customer profile.
This information was crucial in giving a strategic advantage over other retailers, having a massive impact on the company’s performance.
Versatility in scope and application
These are just three examples showing the versatility and power of using prescriptive analytics solutions in your business.
In our next post, we will further demystify Decision Optimization by showing you how to identify Optimization applications. Whether you’re in media, finance, aviation, retail, or hospitality, Decision Optimization ultimately helps you make better business decisions that can put you ahead of your competitors, satisfy more customers and employees, and improve your bottom line.
Want to chat more about specific Decision Optimization solutions for your organization? Schedule a discovery call with our team. And subscribe to our newsletter get our third and final post on prescriptive analytics as soon as it’s ready!
Need help with TM1?
We're here for you