Trust In Autonomous Vehicles

Problem Statement

Several recent research studies have suggested that “lack of trust” is one of the major factors which is slowing the commercial introduction of autonomous vehicles. While the technologies themselves are evolving rapidly and are achieving great leaps forward in driving safety, the human occupants of autonomous vehicles can become uncomfortable due to the lack of unnaturalness and intuitiveness of the situation. Autonomous systems have few precedents in human history and are thus generally unfamiliar to people. In addition, current forms of automation normally communicate information about “state” rather than “intention” during operation, making it difficult for people to fully understand the trajectory of events and nearly impossible to perform predictions regarding what will happen next.

Humans are social creatures who live in a world characterised by verbal and non-verbal interaction with other living creatures. Humans are able to estimate the intentions of other humans or of other forms of life such as animals, but are currently provided little or no evidence of intentionality by today’s highly automated systems such as autonomous vehicles.

The proposed research will investigate the psychological and sociological construct of “trust”. Particular attention will be paid to how humans develop trust in other living creatures and in automation. The information, myths and rituals which help people to develop trust will be identified, and strategies will be defined for the autonomous vehicle to “become a friend” of its human occupants and to gain their trust.

Program Of Work

• Literature review of the most widely accepted psychological and sociological models of “trust”.
• Literature review of trust issues and trust research in relation to any aspect of road vehicle operation.
• Ethnographic activity (interviews + workshops) to extract examples of "trust failures" during the operation of existing road vehicles.
• Ethnographic activity (real fictions) to extract examples of "trust failures" during the operation of future autonomous road vehicles.
• Definition of a "trust framework" which provides a checklist or metric for rapidly evaluating new autonomous vehicle systems designs for "trustworthiness".
• Simulator based driving experiments involving "trust failure scenarios" to validate the "trust framework" and to test for sensitivity to gender, age and culture.
• Final reporting.

Value Of The Research Both Commercial And Academic

The rise of automation is proceeding rapidly and the autonomous vehicle, which constitutes a form of artificial life, will soon be a reality on the streets. The research will address the most fundamental of all human concerns in relation to other forms of life either biological or artificial, and that is the question of trust. Without robust and reliable design strategies for achieving trust, the commercial success of autonomous vehicles will prove difficult to achieve. The research will develop criteria for evaluating whether an autonomous vehicle can be considered “trustworthy or not” and will define design guidelines for introducing the criteria into the automotive design process. Recent research has identified the issue of “trust” as being one of the main roadblocks to the timely introduction of autonomous vehicles, thus research to address this fundamental issue would seem urgent.

Brunel offers a number of funding options to research students that help cover the cost of their tuition fees, contribute to living expenses or both. See more information here: https://www.brunel.ac.uk/research/Research-degrees/Research-degree-funding. The UK Government is also offering Doctoral Student Loans for eligible students, and there is some funding available through the Research Councils. Many of our international students benefit from funding provided by their governments or employers. Brunel alumni enjoy tuition fee discounts of 15%.