AbstractThis research seeks to investigate the inherent concerns held by future users of AV by conducting a multi-population survey to obtain how their specific concerns will affect the uptake of AV. An 11-point Likert scale survey instrument with 34 items questions was developed and distributed using different online channels to targeted road users in the UK. The survey population, a total number of 235 people, belong to different demographic segments of road-user population. An initial data processing and analysis was conducted using the SPSS statistical tool to examine the various components of the data based on demography. The pre-analysed data were modelled using machine learning algorithms and fuzzy logic inference tool in MATLAB/Simulink to develop a Fuzzy Logic Autonomous Vehicle Adoption Model (FLAVAM). The data was divided into training and testing sets according to the different categories of concerns held by each user. From the review of literature, safety, trust, privacy, accessibility, and ethics were identified to act as the most predominant concerns that will affect the adoption of AV.
There are several contributions of this research; firstly, the research identified and quantified the impact of diverse causal factors on the adoption of autonomous vehicles and the effect of perceived causal factors on user degree of adoption. Secondly, computational model was developed based on user opinion and perception, which supports effective visualisation of relationship between user adoption and the causal factors under investigation. Thirdly, a custom fuzzy logic model to forecast user adoption of autonomous vehicles which achieved superior performance compared to standard machine learning techniques.
The FLAVAM model provides a new understanding of how inherent/perceived concerns affect the degree of AV adoption autonomous vehicles.
|Date of Award||Nov 2021|
|Supervisor||Vasile Palade (Supervisor), Rahat Iqbal (Supervisor) & Charalampos Karyotis (Supervisor)|
- autonomous vehicles
- technology acceptance and adoption
- fuzzy logic
- machine learning