2024-03-29T11:39:14Zhttps://www.tdx.cat/oai/requestoai:www.tdx.cat:10803/6688232020-03-19T16:11:01Zcom_10803_183col_10803_234
nam a 5i 4500
Driving complexity
Adaptive HMI
Machine learning
Design principles
Infotainmen
Proposal of an adaptive infotainment system depending on driving scenario complexity
[Barcelona] :
Universitat Politècnica de Catalunya,
2020
Accés lliure
http://hdl.handle.net/10803/668823
cr |||||||||||
AAMMDDs2020 sp ||||fsm||||0|| 0 eng|c
Galarza Osio, Miguel Ángel,
autor
1 recurs en línia (187 pàgines)
Tesi en modalitat Doctorat industrial
Tesi
Doctorat
Universitat Politècnica de Catalunya. Departament d'Enginyeria Telemàtica
2020
Universitat Politècnica de Catalunya. Departament d'Enginyeria Telemàtica
Tesis i dissertacions electròniques
Paradells Aspas, Josep,
supervisor acadèmic
TDX
The PhD research project is framed within the plan of industrial doctorates of the “Generalitat de Catalunya”. During the investigation, most of the work was carried out at the facilities of the vehicle manufacturer SEAT, specifically at the information and entertainment (infotainment) department. In the same way, there was a continuous cooperation with the telematics department of the UPC.
The main objective of the project consisted in the design and validation of an adaptive infotainment system dependent on the driving complexity. The system was created with the purpose of increasing driver’ experience while guaranteeing a proper level of road safety. Given the increasing number of application and services available in current infotainment systems, it becomes necessary to devise a system capable of balancing these two counterparts. The most relevant parameters that can be used for balancing these metrics while driving are: type of services offered, interfaces available for interacting with the services, the complexity of driving and the profile of the driver.
The present study can be divided into two main development phases, each phase had as outcome a real physical block that came to be part of the final system. The final system was integrated in a vehicle and validated in real driving conditions.
The first phase consisted in the creation of a model capable of estimating the driving complexity based on a set of variables related to driving. The model was built by employing machine learning methods and the dataset necessary to create it was collected from several driving routes carried out by different participants. This phase allowed to create a model capable of estimating, with a satisfactory accuracy, the complexity of the road using easily extractable variables in any modern vehicle. This approach simplify the implementation of this algorithm in current vehicles.
The second phase consisted in the classification of a set of principles that allow the design of the adaptive infotainment system based on the complexity of the road. These principles are defined based on previous researches undertaken in the field of usability and user experience of graphical interfaces. According to these of principles, a real adaptive infotainment system with the most commonly used functionalities; navigation, radio and media was designed and integrated in a real vehicle. The developed system was able to adapt the presentation of the content according to the estimation of the driving complexity given by the block developed in phase one. The adaptive system was validated in real driving scenarios by several participants and results showed a high level of acceptance and satisfaction towards this adaptive infotainment.
As a starting point for future research, a proof of concept was carried out to integrate new interfaces into a vehicle. The interface used as reference was a Head Mounted screen that offered redundant information in relation to the instrument cluster. Tests with participants served to understand how users perceive the introduction of new technologies and how objective benefits could be blurred by initial biases.
p
ES-BaCBU
cat
rda
ES-BaCBU
text
txt
rdacontent
informàtic
c
rdamedia
recurs en línia
cr
rdacarrier