Advanced

Developing a strategic controller with haptic and audio feedback for autonomous driving

Guijarro, Sonia and Andjelkovic, Alexander (2015)
Department of Automatic Control
Abstract
Traffic accidents cause over 1.2 million deaths, and tens of millions of people are injured or disabled every year. Advanced driver assistant systems and other safety features have the possibility to reduce traffic accidents but do not account for human errors. Studies show that over 90% of all traffic accidents are caused by human errors. One way to reduce human errors is to introduce automation, and several major car manufacturers predict that autonomous vehicles will be available on the consumer marker as early as 2020. In theory automated cars could reduce deaths and injuries caused by traffic accidents, but there are several issues which need to be solved before it can be realized. One of these issues is how to keep the driver in the... (More)
Traffic accidents cause over 1.2 million deaths, and tens of millions of people are injured or disabled every year. Advanced driver assistant systems and other safety features have the possibility to reduce traffic accidents but do not account for human errors. Studies show that over 90% of all traffic accidents are caused by human errors. One way to reduce human errors is to introduce automation, and several major car manufacturers predict that autonomous vehicles will be available on the consumer marker as early as 2020. In theory automated cars could reduce deaths and injuries caused by traffic accidents, but there are several issues which need to be solved before it can be realized. One of these issues is how to keep the driver in the loop while the car is in autonomous mode. A human-machine interface of a strategic controller for autonomous driving was developed.
Multimodal feedback consisting of auditory and haptic signals was developed for the strategic controller using an iterative design process. A user study was carried out in order to evaluate the multimodal feedback and identify usability issues, and a simulator study was carried out in order to benchmark the concept’s usability.
The strategic controller prototype developed in this thesis allows the driver to take part of the driving process and control of the car by inputting commands. The controller also provides the driver with multimodal feedback based on an analysis of mock-up sensor/image data from the vehicle. User input is either denied or accepted depending on the analysed data, and on demand feedback is also provided related to the general state of the autonomous system.
Multimodal feedback was found to be promising for communicating complex information in humanmachine interactions. Although users had little to no experience of autonomous driving, they found the developed concept to be attractive and would use it for daily commuting. As it is difficult to mirror reality in simulators, test subjects may have had a more positive attitude towards the concept.
However, the issue of keeping the user in the loop still persists. Feedback needs to be designed thoroughly and should not be limited to two modalities. Instead, information should be distributed through several modalities in order to reduce cognitive load and increase the user’s situational awareness. The benchmark of the developed concept showed promising results, although the results may have suffered due to hardware limitations. (Less)
Please use this url to cite or link to this publication:
author
Guijarro, Sonia and Andjelkovic, Alexander
supervisor
organization
year
type
H3 - Professional qualifications (4 Years - )
subject
ISSN
0280-5316
other publication id
ISRN LUTFD2/TFRT--5969--SE
language
English
id
7756450
date added to LUP
2015-08-10 09:19:00
date last changed
2015-08-10 09:19:00
@misc{7756450,
  abstract     = {Traffic accidents cause over 1.2 million deaths, and tens of millions of people are injured or disabled every year. Advanced driver assistant systems and other safety features have the possibility to reduce traffic accidents but do not account for human errors. Studies show that over 90% of all traffic accidents are caused by human errors. One way to reduce human errors is to introduce automation, and several major car manufacturers predict that autonomous vehicles will be available on the consumer marker as early as 2020. In theory automated cars could reduce deaths and injuries caused by traffic accidents, but there are several issues which need to be solved before it can be realized. One of these issues is how to keep the driver in the loop while the car is in autonomous mode. A human-machine interface of a strategic controller for autonomous driving was developed.
Multimodal feedback consisting of auditory and haptic signals was developed for the strategic controller using an iterative design process. A user study was carried out in order to evaluate the multimodal feedback and identify usability issues, and a simulator study was carried out in order to benchmark the concept’s usability.
The strategic controller prototype developed in this thesis allows the driver to take part of the driving process and control of the car by inputting commands. The controller also provides the driver with multimodal feedback based on an analysis of mock-up sensor/image data from the vehicle. User input is either denied or accepted depending on the analysed data, and on demand feedback is also provided related to the general state of the autonomous system.
Multimodal feedback was found to be promising for communicating complex information in humanmachine interactions. Although users had little to no experience of autonomous driving, they found the developed concept to be attractive and would use it for daily commuting. As it is difficult to mirror reality in simulators, test subjects may have had a more positive attitude towards the concept.
However, the issue of keeping the user in the loop still persists. Feedback needs to be designed thoroughly and should not be limited to two modalities. Instead, information should be distributed through several modalities in order to reduce cognitive load and increase the user’s situational awareness. The benchmark of the developed concept showed promising results, although the results may have suffered due to hardware limitations.},
  author       = {Guijarro, Sonia and Andjelkovic, Alexander},
  issn         = {0280-5316},
  language     = {eng},
  note         = {Student Paper},
  title        = {Developing a strategic controller with haptic and audio feedback for autonomous driving},
  year         = {2015},
}