Road Assistant
From c,mm,n
The following text is the summary of a design report by Jos Thalen. He did research in february/may 2007 about a road assistant for the car of the future at the University of Twente.
Abstract
As part of the 'Car of the Future' project, the University of Twente is working on the implementation of 'Advanced Driver Assistance Systems' (ADAS). In 2006 the design of an integrated ADAS system and four user interface concepts was finished. One of the interface concepts was the Road Assistant.
The Road Assistant tries to interact with the driver by using human interaction methods (including for example facial expressions) to reach the driver on an emotional level. For example, if the driver is speeding, the Road Assistant looks angry or scared. This may result in a faster and more reliable and/or effective way of interfacing with the driver.
To investigate the feasibility of this concept, this report starts with a review of relevant literature. Several fields of science are involved in investigating human-machine interactions, affective interactions, and the use of intelligent or conversational agents. Research from these fields of science not only support the Road Assistant, but also provide knowledge and technology to further design the concept.
The first step towards a detailed design is the development of a functional layout. The main functions of the Road Assistant are defined, based on which function specifications and interactions are determined. This results in a functional layout of the Road Assistant, which can be used as a framework for a feasible implementation of the concept.
The implementation consists of finding technological solutions to fulfil the required functions. The reports covers both the exterior design and the interior design of the Road Assistant. A software structure that supports intelligent behaviour. emotional behaviour and adaptive behaviour is proposed. Furthermore, a detailed design of the face, neck and body is presented. The face supports the expression of several emotions, using a minimal amount of moving parts.
The report ends with a first evaluation of the concept based on system requirements. Results of this evaluation are used to phrase recommendations. The recommendations indicate which tests are required to further develop the concept. A test programme has been proposed, focussing on testing the ability to map emotions to traffic situations, and a more realistic usability test using a driving simulator and a functioning prototype of the Road Assistant.
