Bachelor and Master Theses

Background: In order to improve efficiency of saw chains it is necessary to understand the complex and dynamic interaction among the chain, the chainsaw and wood. The components of a saw chain go through a large series of tests which have been optimized to improve the product quality. In one important test, the energy consumption for cutting in wood under well-defined laboratory conditions is measured. The wood itself, however, is an inhomogeneous material. Therefore, energy consumption varies as the chain cuts through the wood.

Wood cutting properties for the chains is measured in the lab by analyzing the force, the torque, the consumed power and other aspects of the chain as it cuts through the wood log. One of the essential properties of the chains is the cutting efficiency which is the measured cutting surface per the power used for cutting per the time unit. These data are not available beforehand and therefore, cutting efficiency cannot be measured before performing the cut.

Cutting efficiency is related to the relative hardness of the wood which means that it is affected by the existence of knots (hard structure areas) and cracks (no material areas). The actual situation is that all the cuts with knots and cracks are eliminated and just the clean cuts are used, therefore estimating the relative wood hardness by identifying the knots and cracks beforehand can significantly help to automate the process of testing the chain properties, saving time and material and give a better understanding of cutting wood logs to improve chains quality.

In a previous MSc project1, we have applied deep learning-based instance image segmentation (in TensorFlow ecosystem) such as Mask-RCNN to detect and segment knots and cracks on an end face of the log. Methods were also developed to estimate pith’s vertical position from the wood image and regenerate equivalent cutting efficiency graph based on knots and crack’s percentage at each vertical position of the wood image.

Scope: The scope of this thesis project to set up and propose a “Digital Twin” that can be used to monitor an entire chain of the assembly of the Power Transfer Unit (PTU). The Digital Twin should be able to capture properties, measurements, and process steps of the PTUs with the help of Machine Learning and AI.

Goal: Major objectives are:

·       Describe how the energy consumption varies as the cut progresses through the wood by conducting image analysis (also open to other non-imaging methods) of the cut surface. 

·       Improve the accuracy and efficiency of deep learning models by experimenting and iterating with various semantic and instance segmentations methods (eg.,in Pytorch ecosystems such as Deeplabv3, Detectron2).

Task Description

The thesis will start with a thorough introduction into how saw chains work and how they are characterized and tested in our laboratory and about the previous work including published MSc thesis.  You shall then investigate different types of existing and potential input data and compile a first list of possible data driven approaches.

Husqvarna’s colleagues in the laboratory will support you in data acquisition, but you are expected to also spend your own time. In the final part, you evaluate the most promising approach on the collected data.

Please send in your application with CV and cover letter no later than November 27th

Machine learning, Data Science, Engineering or Natural Science with a focus on data-driven methods. Good programming skills (eg. Python or C++) and previous experience in software engineering, and machine learning is highly desirable.

The deadline is on 27th November, please contact me as soon as possible with your CV.

Husqvarna: Shaping great experiences is what we do. With our passion for innovation, we create new solutions to enhance urban and green spaces used and loved by many. Transforming the way the world care for outdoor environments. For more than three centuries we have kept innovating and re-inventing our business. Today we’re the world’s leading producer of outdoor power products for forest, park and garden care, watering products and power tools for construction.

https://www.husqvarnagroup.com/en/join-us