At any given time, the Driver Interface Group has anywhere from 2-4 projects underway at any given time. These projects can vary in size, with $100,000 to $500,000 being fairly typical, with durations being from 6 months to 2 years, but there have been multi-phase projects lasting 4 years.
Development and Testing of Scenarios for the GM DriveSafety Simulator
This project will develop and validate a series of software libraries for the GM DriveSafety simulator GM can use for routine tests of the ease of use of driver interfaces (e.g., climate control) and the effectiveness of warning systems. The goal is to accelerate human factors tests at GM by providing allowing for the rapid development of test scenarios that can be used for experiments. The experimenter will have a point and click interface that will allow them to specify the behavior of nearby vehicles and other characteristics, changing scenario development from 1-2 months of writing TCL code to a task that can be easily completed in a week, and sometimes in a few days.
This initial effort will focus on the development of scenarios for expressway
driving and urban driving. In addition, a mini-scenario will be developed
that will be similar to the black lake handling course at the Milford
Proving Grounds. The mini-scenario should be useful for calibrating the
Integrated Vehicle-Based Safety Systems (IVBSS)
IVBSS combines forward collision warning, lane departure warning, curve speed warning, and lane change/merge warning into a single integrated warning system. This project consists of 2 phases, with the first phase (in which the driver interface group was involved) lasting 2 years and involving 4 major simulator experiments and several small ones. Those experiments considered (1) if warnings should be combined (for example all forward warnings to the same sound), (2) what the timing of warnings in succession be (for example if they should overlap or the second should be delayed), (3) how warning delay and accuracy trade off, (4) the timing between sounds for warnings that involve repetition (such as the simulated rumble strip), and other issues. The results of these experiments influence the design of a warning system now begin evaluated in a major on-the-road experiment.
Human Factors and Speech Interfaces
The purpose of this project is to aid the development and evaluation of speech interfaces for navigation destination entry and music selection. The end result was a Monte-Carlo simulation of tasks associated with those systems. To populate those models, data were collected on entry methods used, errors and their probabilities, and task times.
Demonstration and Evaluation of Technologies for SAfety VEhicle(s) using adaptive Interface Technology (SAVE-IT)
The purpose of the project was to examine distraction-related crashes and to explore a variety of means related to the driver interface that might reduce them. In the first phase, literature reviews were conducted as well as both simulator and on-the-road experiments to assess workload in a variety of situations.
In the second phase, the Advanced Collision Avoidance System (ACAS) data
was examined to determine the frequency of various visual, auditory, cognitive,
and psychomotor demands of secondary tasks. In addition, using a rating
method, the subjective workload of a wide variety of driving situations
was assessed, and an equation was developed to reliably compute workload
from in-vehicle sensors in real time.
Human Factors of Intersection Warnings
The purpose of this project is to help Nissan develop and test a driver interface to inform and subsequently warn drivers when they might be at risk of passing through an intersection improperly. Both a literature review and a driving simulator experiment were conducted.