Doing The Math On Logistical Challenges

The application of customised algorithms to difficult logistical challenges is one of today’s most exciting business trends.

While most small to medium-sized companies are still likely to be employing a rule-of-thumb approach and trying to learn from past experiences, larger organisations are increasingly opting to develop their own mathematical models or adapt existing ones that work for their particular needs. Used correctly, it’s clear this cutting edge strategy can yield significant cost savings.

So, what types of specific problems can algorithms help solve? My own modeling research is based on examining production problems from the perspective of a manufacturer who’s trying to optimise a performance measure constituting a combination of different costs. The aim is to try and minimise total costs involved, which are reflected by the time required not only to produce, but also to deliver.

With this in mind, consider a situation where:

• Production takes place at several locations and is performed at different speeds depending on the available technology at each individual facility.
• The use of limited capacity vehicles for moving finished products from these facilities to the customer is causing transportation constraints.
• The goods supply chain routinely encounters a bottleneck. For example, exporting through a seaport will see all goods bottleneck at the dock before proceeding further.
• The ship transporting the product to market has a deadline for leaving the port.

In a new paper we’ve been working on, we’ve actually developed a mathematical model for addressing this very set of problems. While we haven’t yet had the opportunity to test it within the parameters of an actual business environment, my past experience of applying algorithms to real-life situations leads me to believe that, in a manufacturing scenario such as the one stipulated above, cost savings of 15% to 20% could definitely be achievable. Now that makes perfect business sense in anyone’s language.

Author: Associate Professor Daniel Oron -
 Discipline of Business Analytics, University of Sydney Business School