Application of further development of the Forecast Algorithm and reinforcement Learning for effective Sales & Operations planning – Ewals Cargo Care

 Overview

 Company Description

Ewals Cargo Care is a family owned transportation company, which is founded in 1906 by Alfons Ewals. The main office is found in Tegelen near Venlo. Besides this, Ewals has 32 other offices in Europe, which lead to a local presence in 15 European countries with over 2500 employees.

Ewals has grown into a strong international player. Our customers are offered a broad spectrum of logistics products ranging from full- and part loads to Control Tower services. Especially, but certainly not exclusively, we are known  for our European multimodal network that includes more than 4500 Mega Huckepack XL(S) trailers. In addition, we can rely on a solid network of partners. Ewals is also continuously improving work processes and investing in the development of our employees.

Project Description

Project description:

The project’s aim is to improve the process of Forecasting and strategic product planning. Especially in the dynamic and fast-moving world – where disruptive events likely to take place with increasing frequency – it is of vital importance to empower business planning in foresight and increase business predictability. The project touches both the commercial and operations domain – yielding effective Sales & Operations planning.

The project is organized within the Product Intelligence department – in which multiple TU/e alumni reside, whom focus on research and development and technological possibilities to enact upon, to become ready for the future.

In collaboration with the European Supply Chain Form, Ewals Cargo Care and the ESCF have composed the “UNIVERS” Roadmap, which is journey in joint effort, supporting and enabling Ewals with the effective transition to a data driven business model. In an earlier phase of the roadmap an Forecast model has been created (where we internally refer to as the “Steering Module”) – in this context you are required to further develop a comprehensive V2.0 for Ewals Cargo Care that encompasses all the critical aspects of effective Forecasting.

Reinforcement learning (RL), a type of machine learning where an agent learns to make decisions by interacting with its environment, can potentially be used to optimize the current Forecast algorithm. It operates on the concept of reward optimization – an agent, through trial and error, learns to make decisions that maximize a cumulative reward. In this scenario, the ‘reward’ can be defined based on multiple parameters including strategic aim of Ewals Cargo Care (resilience and diversification strategy), competitiveness, maximized profitability, effective capacity utilization and more. One does logically not exclude the other.

Goals of the project:

Your primary task for this project will involve further developing a reinforcement learning model – improving the current model already existing and in use – to improve/ optimize the accuracy of the Forecasting model. The RL model should: (1) leverage the historical data and strengthen the current algorithm with new value added parameters (e.g., macro level variables) (2) reflect strategic development direction of Ewals Cargo Care to propose decisions how and where to strategically develop (and grow).

Deliverables:

• A comprehensive RL prediction model (i.e., Forecasting 2.0).
• A report detailing the development process, the challenges faced, the solutions implemented, and the final results. This report should include an analysis of the effectiveness of the proposed system in optimizing the Forecasting process.

Essential student knowledge:

This project provides an exciting opportunity to delve into the intersection of Sales & Operations planning and reinforcement learning for optimizing logistics operations. It will not only equip you with valuable experience in these advanced technologies but also contribute meaningfully to a more sustainable and efficient logistics industry. As a master student, you are expected to demonstrate a clear understanding of the techniques employed, and your ability to creatively and effectively apply these techniques to real-world challenges. You are encouraged to take a proactive approach in problem-solving, consider alternative methods where necessary, and communicate your findings clearly and concisely.

• Programming knowledge (preferential Python, SQL)
• Reinforcement Learning Modelling
• Supply chain knowledge and affinity (Transport and Logistics)

 More information: escf@tue.nl