Page Title Here

Below is a list of all the blog posts you are posting that your
visitors might be interested in...

< Go back

Reference case

Process Optimization: Filling and Shipping containers with e-llis

Bram Vanschoenwinkel
Bram Vanschoenwinkel

is an AI expert who started his career as a researcher at the Computational Modelling Lab at the University of Brussels and obtained a Ph.D. in Science (Machine Learning, Data Mining). Over the past 10 years, Bram led a team of 15 data scientists offering actionable business insights from data. He is currently Chief Product at Customaite.

Process optimization is a must for companies that want to maintain their competitive advantage – something logistics operator e-llis knows all too well. Frequently finding itself forced to occasionally ship its customers' containers which are only forty percent filled, the company called on us for help. In co-creation with e-llis, our team worked on a proof of concept to optimize the shipping process. The ambitious goal: to increase the filling level of these containers to eighty percent.

Defining the status quo

To enable ourselves properly to carry out the necessary optimizations, we set out to define the status quo. Instead of relying on gut feeling, we used concrete figures to answer the following questions: what is e-llis’ performance like today? How often does e-llis ship containers for its customers when they are only ten to forty percent full?

Process optimization based on 3 scenarios

The figures e-llis shared with us soon indicated that there was indeed room for improvement in the margins concerning the containers’ filling levels. We distinguished three different scenarios e-llis tends to face on an everyday basis.


Operational Excellence E-llis


Containers filled to the brim

Ideally, a container is fully loaded. E-llis fills it to the brim and takes it to the port, where it begins its journey, which typically starts in Asia and ends in Europe. Upon arrival, the container is removed from the ship and stored in a European warehouse. E-llis fills the majority of the containers in this way. This solution is obviously the simplest and most cost-efficient.

Pricey mandate

But what if the load is much smaller than the container? In that case, e-llis offers two options – the first of which is the most expensive. In a nutshell, the ocean freight partner gets a mandate to fill the remaining cubic meters with other loads. The only requirement is that the cargo must reach the right (European) port. Once there, the cargo is picked up and taken to the warehouse. However, this option is so pricey that most companies decide to ship containers themselves once they’re 35 percent full.

Making the little things count

The second and final option for e-llis is to fill every cubic meter available by combining small orders. It is this option the company is determined to optimize in co-creation with us, more specifically by bundling various suppliers to the same port. An important precondition is, however, that each order must have the same port of departure and arrival. Ideally, the orders even have the same destination to reduce costs even further.

Dashboard simulation in co-creation

Following an extensive feasibility analysis, we started to develop a dashboard simulation based on a smart algorithm, enabling e-llis to gain more insight into the incoming orders. It also became immediately clear how e-llis could optimize their handling of the orders in question – with special attention to the filling levels.

The algorithm calculates how loads can be optimally distributed over different containers, so that as few cubic meters as possible remain empty. Data manipulation takes place in the backend in R, with Shiny as front-end to maintain a user-friendly overview.

Next stop: operational excellence

With their new tool in hand, e-llis can now easily explore different scenarios to optimize the filling level of containers. The business value of this project therefore lies mainly in the insights it generates, which in the long term will facilitate a substantial cost reduction and will bring e-llis another step closer to achieving operational excellence.


Curious how we can help you build a user-friendly digital assistant? Fill in the form below and we'll contact you asap.