Background

This thesis is suitable for an engineering student in mechanical engineering, electrical engineering, computer science or equivalent with an interest in Data Science/AI with a strong interest in mechanical systems, software coding and virtual simulation. The desired start date is Feb 2026.

Modern construction-equipment drivetrains are becoming increasingly complex due to the introduction of electrified powertrains, advanced transmissions, torsional dampers, clutches, and other dynamic components. As a result, torsional vibration can originate from several interacting sources, including engine excitation, electric-machine torque ripple, gear meshing, torsional stiffness, damping, inertia, and operating conditions.

Identifying the root cause of an observed torsional vibration response is often challenging. Measurements can reveal that an abnormal vibration exists, but determining which physical parameter or component is responsible may require extensive testing and engineering analysis.

Numerical simulation provides an opportunity to investigate these relationships systematically. A validated driveline dynamic model can be used to generate large numbers of controlled simulation cases by varying physical parameters and operating conditions. These simulation data can subsequently be used to develop AI/ML models capable of identifying the most likely root cause of an observed vibration behaviour.

This work mainly takes place at the Division of Powertrain,Technical Center in Eskilstuna.

Project summary

This thesis offers the opportunity to work at the intersection of:

Automotive Engineering × Noise & Vibration × Numerical Simulation × AI

The project is particularly suitable for a student who wants to understand both the physical engineering problem and the AI methods used to solve it.

The work has potential for further development toward AI-assisted NVH diagnostics and simulation-driven engineering applications.

The final outcome will be a validated methodology demonstrating how a physics-based driveline simulation model can be used to generate training data for AI-assisted identification of torsional vibration root causes. The thesis will combine engineering understanding, numerical simulation and modern AI techniques in a practical industrially relevant problem.

Suitable background

You should have:

• Strong analytical and problem-solving skills
• Interest in vehicle/driveline dynamics and Noise & Vibration
• Basic knowledge of numerical modelling
• Programming experience in Python and/or MATLAB
• Interest in machine learning and data-driven engineering

Experience with 1D simulation, vibration analysis, signal processing, CAE or machine learning is an advantage.

The successful candidate should be willing to work at the Division of Powertrain department,Technical Center in Eskilstuna. Last day to apply is November 16. 

Contacts

The supervisor for this thesis project is:

Ramkumar Kandasamy

Powertrain | Transmission Development,

Volvo Construction Equipment,

CE40960, TCC Building,

SE-631 85 Eskilstuna, Sweden.

Telephone: +46 (0)16 201 9984

Mobile: +46 769409984

E-mail: ramkumar.kandasamy@volvo.com