Autonomous Drive User Interface
Project type: Interface Design / Editorial Design
|This research analyzes the current scenario and future implications of autonomous driving systems, addressing the creation process of an interface able to communicate intentions and movements of autonomous vehicle.||La tesi analizza lo scenario attuale e le implicazioni future dei sistemi di guida automatizzata, affrontando il processo di creazione di un’interfaccia atta a comunicare le intenzioni e gli spostamenti del veicolo autonomo|
Date: Oct 2013 – Jun 2014
MA Thesis Project
This research analyzes the current scenario and future implications of autonomous driving systems, addressing the creation process of an interface able to communicate intentions and movements of autonomous vehicle (human intelligence and artificial intelligence should never come in conflict, but constantly communicate). The ultimate goal is to increase the confidence felt by the user to this system, encourage the inclusion, acceptance and all the benefits it could bring to mobility (final research book is intended as a publicizing tool useful to understand the complexity and the solid architecture of autonomous vehicles). After a survey of prototypes, technologies, existing scientific publications and interfaces for automotive industry, I carried out an analysis about the actual UI designing process, prototyping and testing.
Current studies in automotive field shows an imminent transformation of cars in intelligent and interconnected devices: one of the major revolutions of the century in automotive field that has the potential to change private and individual mobility. It envisages a scenario in which car is not only a tool to get from point A to point B, but a complex device that receives and provides information, is connected to the network, knows traffic situation, programs maintenance, can recognize emotional states, hypothetical distractions and driver physical problems (sleep or tiredness), is able to teach the correct way to drive through feedback and, vice-versa, learning input from the driver: it may ultimately drive independently without human help. This UI is designed for the first models of self-driving cars with manual drive option and standard cockpit layout.
In this scenario car social role is in clear evolution, such as the role of the driver itself. The crucial point concerns designing instruments from engineering and user experience point of view, in other words in which way the human and the machine will interact. GUI is an essential tool to create a bridge between this two intelligent entities in order to allow them to communicate harmoniously. During the graphical interfaces research phase, it has been crucial to determine visual variables complexity, information hierarchy and typology, and cognitive overload during manual driving mode.
The first section of the thesis explores the general discourse on cars, the social role they acquired over the years, advantages and disadvantages on society and mobility. Regarding Autonomous Drive I analysed current players which compete in autonomous vehicles design, technologies used, how the system works, and mobility changes that autonomous drive systems and connected cars could produce.
The second part introduces user interface design which makes explicit the objectives of the research: to build a bridge between the human and the machine and increase the confidence in autonomous system decisions. “Cleo” is the name of the designed system. It’s a ‘visual storyboard’ of what it’s happening outside the car. It shows Autonomous Drive immediate future movements and choices. “Why is the car turning right?”, “Which way is it taking?”, “Can the car detect the cyclist in front of me?”. After a first prototyping stage I explain UI through a deep designing choices analysis.
The interface is based on two fundamental principles: simplicity (of use) and anthropomorphism (of language). It’s not designed for an audience of engineers or designers, it is designed to be as simple as possible, usable by anyone. Even in this case I initially carried out a survey on current vehicles user interfaces, methods of interaction, common problems, etc.
One of the most important aspects on which I worked is to maintain internal consistency of the vehicle (interface design and interior design [dashboard] must not be disconnected). I designed various navigation layouts with different information arrangements which were subsequently analysed by a representative sample of users in order to know which layout was better (in which layout was simple and straightforward to find required information).
The next step was a user test by using an eye tracker. I developed a rapid prototype of the interface, then I asked a small sample of users (7 people) to find specific informations and activate specific functionalities on a digital platform to understand which operations were more difficult. It is not a scientific test, however it brought out some design problems. It was not possible to perform a scientific test for a lack of instrumentation (I used “Pupil Dev” by Pupil Labs running on Ubuntu, showing the prototype interface on an iPad). Some interface features have been modified according to the problems identified by users. For example, test showed some ambiguity between the system settings button and map settings button (resolved by changing the icon type, the background contrast and the position on the screen), I also simplified the procedure to change the layout by eliminating the activation confirmation (users expected an immediate change of layout), and solved other usability issues. No user found difficult to understand the behavior of the car.
Physical media designed to accommodate the interface are different (a screen placed on the dashboard, an instrument panel in front of the driver, and a HUD) and are thought for autonomous driving functionalities but also, of course, for a traditional manual driving so that those same media used in manual driving should not distract the driver. In order to do this large distracting display surfaces have been avoided and different information arrangements and density between manual driving and autonomous driving have been chosen.
As final step some of the features designed for onboard interface were carried on mobile devices in order to manage the car remotely, planning routes, maintenance, routines.
It was also suggested a system of autonomous taxi / autonomous car-sharing through which users have the chance to request a car on-demand anytime they want, avoiding the need to own a car at all.