Personalized Software in Heavy-Duty Vehicles - Exploring the Feasibility of Self-Adapting Smart Cruise Control Using Machine Learning

Aug 5, 2026

Upload by Aria

This study explores feasible ways to personalize driving functions for heavy-duty vehicles. It uses machine learning to predict vehicle state velocity, independent velocity, and to classify driver behavior, combining Long Short-Term Memory (LSTM) neural networks with traditional classification methods. Data were collected from the vehicle using CAN bus signals and map-based data, then preprocessed for training and evaluation. Results are satisfactory across models, with one-second velocity predictions outperforming ten-second predictions, and classification accuracy reaching 93%–99%. The findings highlight both promise and limitations of LSTM and classification approaches, and motivate future work on transfer and deep learning with more suitable data.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages89

Reading Time3 h 44 min

AccessFree to download

Abstract

Keywords

Acknowledgements

List of Acronyms

Personalized Software in Heavy-Duty Vehicles Exploring the Feasibility of Self-Adapting Smart Cruise Control Using Machine Learning Master’s thesis in Complex Adaptive Systems and Systems, Control and Mechatronics Charlotte De Geer & Alex Matsson DEPARTMENT OF ELECTRICAL ENGINEERING CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2023 [www.chalmers.se](www.chalmers.se) Master’s thesis 2023 Personalized Software in Heavy-Duty Vehicles Exploring the Feasibility of Self-Adapting Smart Cruise Control Using Machine Learning Charlotte De Geer & Alex Matsson Department of Electrical Engineering Systems and Control Chalmers University of Technology Gothenburg, Sweden 2023 Personalized Software in Heavy-Duty Vehicles Exploring the Feasibility of Self-Adapting Smart Cruise Control Using Machine Learning Charlotte De Geer & Alex Matsson © Charlotte De Geer & Alex Matsson, 2023 . Supervisor: Robin Karlsson, Volvo GTT Examiner: Jonas Fredriksson, Chalmers Department of Electrical Engineering Master’s Thesis 2023 Department of Electrical Engineering System and Control Chalmers University of Technology SE-412 96 Gothenburg Telephone +46 31 772 1000 Typeset in LATEX Printed by Chalmers Reproservice Gothenburg, Sweden 2023 Personalized Software in Heavy-Duty Vehicles Exploring the Feasibility of Self-Adapting Smart Cruise Control Using Machine Learning Charlotte De Geer & Alex Matsson Department of Electrical Engineering Chalmers University of Technology Abstract This study aims to explore possible and feasible ways to personalize driving functions for heavy-duty vehicles. The idea is to use machine learning algorithms, specifically focusing on Long Short-Term Memory (LSTM) neural networks and traditional classification algorithms for current state velocity predictions, independent velocity predictions, and driver classification. The goal is to explore potential approaches for enhancing the existing software to improve the vehicle’s drivability while not compromising fuel consumption. The research methodology involved collecting relevant data from the heavy-duty vehicle, including various readings using the CAN bus and map-based data. The data was preprocessed and used to train and evaluate the LSTM neural network and traditional classification algorithms. The results obtained were satisfactory for all of the models. The predictions from the LSTM models were adequate. The one-second velocity predictions were favorable when compared to the ten-second velocity predictions. From the training progress, it is possible to see that the model learns and identified trends. Furthermore, the classification accuracy using traditional and LSTM classifiers ranged from 93 % to 99 % . These findings highlight the challenges and limitations of employing LSTM neural networks and traditional classification algorithms for software adaptation. Further research is necessary to explore alternative approaches, such as using sufficient and more suitable data for transfer- and deep learning. The insights gained from this study help comprehend machine learning applications in heavy-duty vehicles and suggest future research efforts to enhance software adaptation and thus improve vehicle performance. Keywords: Long Short-Term Memory, LSTM, classification, driver behavior, adaptability, machine learning, ML, neural network, time series. Acknowledgements We would like to express our deepest gratitude and appreciation to our three supervisors, Klara Palm, Oscar Olesen, and Simon Lillskog. Their support, insightful guidance, and constructive feedback have been invaluable in shaping this master’s thesis. We are also grateful for Robin Karlsson, his mentorship, and the opportunities he has provided us during this journey have been essential. We extend our thanks to Jonas Fredriksson for his expertise and critical evaluation of this thesis. Klara, Oscar, Simon, Robin, and Jonas’s thorough examination and valuable

FAQs

What is the main goal of the study on heavy-duty vehicles?

The study aims to explore feasible ways to personalize driving functions for heavy-duty vehicles without harming fuel consumption, while improving drivability.

Which machine learning methods are used in the research?

The research focuses on Long Short-Term Memory (LSTM) neural networks and traditional classification algorithms for velocity prediction and driver classification.

How is the data collected and evaluated?

Relevant vehicle data are collected using CAN bus signals and map-based data, preprocessed, then used to train and evaluate both LSTM and traditional classifiers.

Aug 5, 2026

Upload by Aria