Optimising logistics with AI: Jarro Teunissen’s placement at DSV

Can you tell us a bit about DSV and what they do? 

Jarro: 'DSV is a logistics company, one of the largest in the world at the moment, with more than 160,000 employees. In the Netherlands alone they have over twenty warehouses. They handle the logistics behind the transport of goods, so they arrange the transport side of things for a client whose products are stored in the warehouse. In my section I am mainly involved with the transport of parcels. Think of products such as clothing and electronics, in reality, it covers a wide range, from apparel and consumer goods to medical supplies, batteries, and even motors.' 

What are you working on during your placement? 

Jarro: 'My team deals with what they call event management, which means they look at parcels and pallets that have not been delivered or have arrived late, and try to find out why. When I started I quickly noticed that everything was being done very manually, with a lot of Excel files and a lot of switching between different track and trace websites, because every carrier like DHL, DPD and UPS has its own tracking system. After just one week I thought: why can we not build a tool that does all of this in one place?’  

‘So, I built a web application where you can upload an Excel file and it automatically retrieves all the tracking information for every order using API connections to the carriers. Everything is consolidated in one place, with a dashboard and filter options, while delivery fields and status updates are automatically filled in.’ 

What comes next once the tracking side is in place? 

Jarro: 'The original idea behind the project was to build a model that can predict whether a parcel is going to be delayed before it actually happens. For example, if a parcel has been sitting on a hub for too long, the model should be able to flag that as a likely delay. You could then respond proactively rather than reactively, and maybe decide that a different carrier would perform better in a particular region. Before we can get to that prediction step though, a number of earlier steps needed to be improved first, which is what the tool addresses. The prediction element is still something I want to improve before my placement ends.' 

What knowledge from your studies have you been able to apply? 

Jarro: 'Data engineering has been very directly applicable, building a working front end and back end, using APIs, handling data in a structured way. I am also now applying machine learning for the prediction side of things. But honestly, one of the biggest things has been planning. On the ADS&AI study programme you have a lot of freedom and you have to organise your own time, plan your own weeks, and that is exactly what working in a company requires too. The transition was not difficult for me because I was already used to that way of working.' 

How did you find your work placement? 

Jarro: 'Honestly, finding a placement was harder than I expected. I thought it would be straightforward because so many companies are looking for people with these skills, but the problem is that there are not many experts who have the time to provide proper supervision. In the end, I decided to approach it differently by sharing a post on LinkedIn instead of only applying to vacancies. That turned out to be very effective, as it reached a much wider network. It showed me how important it is to build and use your network, by sharing your work and letting others amplify it, you can create many more opportunities for yourself.' 

What has the experience taught you beyond the technical side? 

Jarro: 'A lot, actually. I have learned a great deal about how a large company works, how to navigate it, and how important it is to just approach people directly rather than waiting. If you build the right connections around you, you get a lot of help and knowledge. Going to talk to the people who actually do the work every day has been really valuable, because that is where you learn how things really function. I did not know how their daily tasks were structured when I arrived, and I still do not know everything, but I have learned a lot more in the past ten weeks than I knew before.' 

What are your plans after this placement? 

Jarro: 'I’d like to explore the field of data science and AI further, for example by moving into computer vision and working with images and other visual data. Healthcare also appeals to me as a sector. I have already done logistics twice now, once in a project in year three and now on placement, so I would like to explore something on the other side of what is possible. For my final year I am thinking about doing another placement.' 

Connect with Jarro Teunissen on LinkedIn or take a look at his work on his portfolio website.