About ORDinL
The ORDinL (Operational Research and Data science in Logistics) project aims to provide companies in the logistics sector with a methodology to improve their decision making processes by applying a data-driven approach.
As Ratliff stated in 2013, "For most supply chain and logistics operations there is an opportunity to reduce costs by 10 to 40 percent by making better decisions". However, despite recent impressive advances in various scientific domains such as machine learning, data mining (DM), operational research (OR) and software development more generally, the ambitions concerning logistics cost reduction remain largely unfulfilled. The essential underpinning paradigms and technologies have continued to evolve independently whereas they should be fully integrated by way of what we call data-driven logistics operational optimisation. Operational optimisation techniques must get the right data at the right time and identify patterns such that they can be exploited alongside real-time data. This way, techniques that were restricted to only simple models can be extended to also handle real-world characteristics.
In this project, several research groups and partners work together to integrate various domains and their techniques such that company data can automatically be translated into information and better decisions. The overall ambition is to develop a software-enabled data-driven logistics methodology that has three unique charachteristics:
1) it is driven by up-to-date data and user feedback,
2) it employs intelligent optimisation technology, and
3) it analyses data and feedback in order to not only steer the optimisation and decision-support but also to continuously improve and adapt itself over time.
To achieve our goal, we will build on and advance the state-of-the-art in the domains of combinatorial optimisation approaches to logistics; optimisation algorithms and problem decompositions; algorithm selection, configuration and construction; data mining and learning techniques; and constraint acquisition and learning.
This project is funded by the FWO as part of the Strategic Basic Research (SBO) programme.
Name: Data-driven logistics
Nr: S007318N
Date: 1/1/2018 - 31/12/2021
Budget: around 2,250,000 EUR
Representatives: Greet Vanden Berghe (KU Leuven), Kris Braekers (UHasselt), Tias Guns (VUB)