30 credits –Statistical forecasting for contract pricing and aftermarket growth i Sodertalje

Scania is currently undergoing an exciting transformation from a heavy truck manufacturer to a provider of sustainable transport solutions. Data driven decision making and business intelligence is a key factor for optimizing sustainable transport solutions. A thesis project at Scania is an excellent way of making contacts for your future working life. Many of our current employees started their career with a thesis project.

Quality Information is a group of analysts and data engineers at Scania working with advanced data analysis and management of information connected to the operation and quality of our products. Our knowledge and competence is used to further improve our customers operation economy hence Scania’s profitability which is directly linked to Scania’s corporate targets.

Background:
The aftermarket, repair and maintenance, is a major source of revenue for Scania. To grow our aftermarket business, we use data science to price repair and maintenance contracts and optimize marketing actions. The Quality Information group supports these efforts with reporting and statistical forecasts of repair and maintenance needs of different markets and vehicles, based on warranty and vehicle operational data.

Project descriptions:
We have openings for several thesis projects aimed at improving the quality, performance, and capabilities of our repair forecasts.

• Optimal pricing populations: current pricing models divides our fleet into vehicle subpopulations with similar characteristics based on market structure and expert knowledge. This project will explore statistical methods to automate and optimize this classification in order to expand the forecasts to new markets and segments.
• Generative repair models: current forecasts are focused on predicting average repair needs of large vehicle populations. This project will formulate and test a fully probabilistic model capable of quantifying prediction uncertainty, model small populations, and merging information from different sources in a principled way.
• Market action forecasts: repair contracts often run for multiple years, and most are sold with new vehicles with no existing repair history. In this project, we will explore modifications of current pricing models to suit the forecasting needs of market actions, which means making shorter forecasts that take into account the pre-existing repair history of individual vehicles.

Education:
We seek motivated students at the M.Sc. level in science, engineering or applied mathematics, with a strong mathematical background, an interest in statistical modelling, machine learning, or data science, and good numerical programming skills.

Number of students: 1-2.

Start date: Late autumn 2018, or Spring 2019.
Estimated time needed: 20 weeks.
Business area: Research and Development

Location: the project will be conducted at Scania in Södertälje, Sweden.

Contact persons:
Martin Lindén (supervisor), YQI, 0700 86 98 20,
Jonatan Elnäs (group manager), YQI, 08-55389763,

Enclose CV, personal letter and relevant transcripts of record. Please high-light your

• knowledge and experience in statistical modelling, machine learning, data science or similar,
• knowledge and experience with R, python, SQL, or similar tools for data science and statistical computing,
• extracurricular activities and leadership experience.

Selections will be made throughout the application period.

• Title: 30 credits –Statistical forecasting for contract pricing and aftermarket growth
• Business area: Research and Development
• Country: Sweden
• City: Sodertalje
• Last application date: 2018-11-30
• Job Id: 20182649