FICO Honors Amazon Team in Competition for Groundbreaking Optimization Research

Rhea-AI Summary

FICO has announced that an Amazon research team received recognition in the 2024 FICO® Xpress Best Paper Award competition for their groundbreaking optimization research. The team's paper, published in the European Journal of Operational Research, demonstrates how to optimize large-scale delivery networks using FICO® Xpress Solver.

The research addresses the challenges of modern online retail delivery networks by developing a sophisticated heuristic approach that can handle thousands of origins and destinations, along with over 100,000 commodities. The solution involves processing approximately half a billion variables and constraints, significantly advancing the field beyond previous research that was to hundreds of origins, destinations, and commodities.

Led by Cristiana L. Lara of Amazon's Modeling and Optimization group, the team utilized FICO® Xpress Solver's powerful optimization engine and API to develop their mixed-integer programming model. This breakthrough has direct applications for businesses operating large distribution networks and builds upon resource-task networks commonly used in chemical engineering process design.

FICO® Xpress Solver used by Modeling and Optimization group at Amazon to perform important research work on improving the speed and efficiency of massive delivery networks

BOZEMAN, Mont.--(BUSINESS WIRE)-- FICO (NYSE: FICO): Global analytics software leader FICO today announced that a research team from Amazon was honored with a distinction in the 2024 FICO® Xpress Best Paper Award competition, acknowledging their outstanding research. Their academic paper presents research on how to optimize an extremely large delivery network, using FICO® Xpress Solver to perform the vast calculations required, involving hundreds of millions of variables.

The paper has been published in the renowned European Journal of Operational Research: https://www.sciencedirect.com/science/article/abs/pii/S0377221723000371

More information: https://community.fico.com/s/blog-post/a5QQi000002rzU1MAI/fico5206

The growth in online retailing has created the need for extremely large, extremely efficient delivery networks that operate around the clock. The Amazon team studied the design of a low-cost fulfilment network that allows for the fast shipping of packages between origins (e.g., warehouses) and destinations (e.g., customer locations).

Service Network Design problems, as they are known, are considered very hard to solve efficiently. Existing research was mostly restricted to cases with a few hundred origins, destinations and commodities. Yet, using the powerful FICO® Xpress Solver, the Amazon team developed a sophisticated heuristic approach that solves for thousands of origins and destinations, and more than 100,000 commodities.

“This kind of problem involves half a billion variables and constraints, and the Amazon team did breakthrough work in this area at this scale,” said Dr. Timo Berthold, director of Mixed-Integer Optimization at FICO who co-manages the contest. “Their work builds on resource-task networks, a modelling framework that is widely used in chemical engineering process design. The Amazon team developed a mixed-integer programming model that is solved by FICO Xpress and has wide applicability for extremely large computational problems with network flow underpinnings.”

“The potent API of Xpress enabled us to tailor our algorithms precisely to our needs, and the powerful optimization engine of FICO Xpress Solver was indispensable in achieving our results,” said Cristiana L. Lara of Amazon’s Modeling and Optimization group, who led the Amazon research team. “Our work has direct application to businesses like ours that rely on massive distribution networks.”

In 2022, the FICO® Xpress Best Paper Award was introduced to acknowledge exceptional research in Mathematical Optimization, Operations Research and related fields. Nominations for the 2025 award are due by February 14.

Recognized as one of the fastest and most powerful optimization solvers, FICO® Xpress Solver gives business users, data scientists and researchers greater power and flexibility when solving complex problems in supply chain optimization, energy, pricing and many other applications.

About FICO

FICO (NYSE: FICO) powers decisions that help people and businesses around the world prosper. Founded in 1956, the company is a pioneer in the use of predictive analytics and data science to improve operational decisions. FICO holds more than 200 US and foreign patents on technologies that increase profitability, customer satisfaction and growth for businesses in financial services, insurance, telecommunications, health care, retail and many other industries. Using FICO solutions, businesses in more than 80 countries do everything from protecting 4 billion payment cards from fraud, to improving financial inclusion, to increasing supply chain resiliency. The FICO® Score, used by 90% of top US lenders, is the standard measure of consumer credit risk in the US and has been made available in over 40 other countries, improving risk management, credit access and transparency. Learn more at www.fico.com.

Learn more at https://www.fico.com/en.

Join the conversation at https://twitter.com/fico & https://www.fico.com/blogs/.

For FICO news and media resources, visit https://www.fico.com/en/newsroom.

FICO is a registered trademark of Fair Isaac Corporation in the U.S. and other countries.

View source version on businesswire.com: https://www.businesswire.com/news/home/20250203043427/en/

Julie Huang

press@fico.com

Source: FICO

FAQ

What achievement did Amazon's research team receive in FICO's 2024 Xpress Best Paper Award competition?

Amazon's research team received a distinction for their research on optimizing large-scale delivery networks using FICO® Xpress Solver.

How does Amazon's optimization research improve upon existing delivery network solutions?

The research handles thousands of origins and destinations with over 100,000 commodities, significantly expanding upon previous solutions that were to hundreds of locations.

What is the scale of variables and constraints handled in Amazon's delivery network optimization model?

The optimization model processes approximately half a billion variables and constraints.

Where was Amazon's optimization research paper published?

The research paper was published in the European Journal of Operational Research.

How does FICO's Xpress Solver contribute to Amazon's delivery network optimization?

FICO Xpress Solver provides the powerful optimization engine and API that enables processing of massive calculations required for the large-scale delivery network optimization.