Federal Highway Research Institute Evaluation of the Tactile Detection Response Task (TDRT) in a laboratory test using a surrogate driving set-up Roland.

Federal Highway Research Institute

Evaluation of the Tactile Detection Response Task (TDRT) in a laboratory test using a surrogate driving set-up

Roland Schindhelm and Eike Schmidt

Outline

1. Characteristics of the TDRT method

2. Background, research questions

3. Method, experimental design

4. Results

5. Conclusions

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 2

Characteristics of the TDRT

TDRT is a method developed for the evaluation of information and control systems HMI (Engström et al., 2005).

TDRT measures effects of secondary task load on driver attention.

Main characteristics of the TDRT:• Stimuli-response task• Tactile stimuli presented by a vibrator

fixed to the participants shoulder• Stimulus duration 1 s, every 3 – 5 s• Manual response by pressing a finger

button

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 3

Background

• 1 tactile mode (TDRT) and 2 visual modes are included in the DRT method.

• Standardization of DRT method in progress (ISO/CD 17488).

• BASt is involved in coordinated international studies which aim at supporting the ISO Task Force in developing the DRT standard.

• Main research questions are related to– the use of DRT in different driving set-ups– difference in results obtained with the three DRT modes– sensitivity and specifity of the DRT modes

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 4

Research questions of the BASt study

The BASt DRT study focuses on the Tactile DRT (TDRT). 

• To what extent is the TDRT sensitive to systematic manipulation of task load for different types of tasks?

• How does the TDRT affect secondary task and primary task performance?

Selection of Research questions from presentations given by Engström and Young, Yokohama Meeting 2012

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 5

Method of the BASt study

Laboratory test performed in July/August 2013

• Subjects: 10 female, 12 male, age 19-64 (M=41.7), licensed drivers

• Single-task and multiple-task scenarios derived from selected combinations of the following independent factors:

– Type of primary task (visual, auditory tracking)– Type of secondary task (N-back Task, SuRT)– Task difficulty (easy, hard)– Use of TDRT (without, with)

• Within-subject design

• Order of task scenarios was counter-balanced across subjects

• Duration per trial: 60 sec.

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 6

Tasks, independent variables

• Driving task surrogate as a primary task:

Continuous Tracking Task (semi-static)Manual control of the cursor viasteering wheel- 2 modalities of feeding back

tracking deviation(visual, auditory)

- 2 difficulty levels (easy, hard)

Easy Hard

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 7

Tasks, independent variables

• Secondary tasks:

SuRTVisual search; manual control of the cursor via key pad- 2 difficulty levels (easy, hard)

N-Back task Auditory stimuli; cognitive processing; vocal response- 2 difficulty levels (easy, hard)

0-Back 1-Back

Easy Hard

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 8

Easy Hard

Experimental set-up

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 9

Dependent variables, indicators

• TDRT:– mean response time– hit rate

• Tracking task:– root mean square deviation

• N-Back task:– percentage of correct answers

• SuRT:– mean response time

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 10

0

100

200

300

400

500

600

700

800

900

1000

Me

an

re

sp

on

se

tim

e

(ms)

p < .01 n.s. n.s.n.s.

Secondary Task

easy hard easy hard

    N-Back SuRT

Primary Task easy hard   easy easy hard

Visual Tracking Aud. Tracking TDRT Tactile DRT

TDRT response time in different task scenarios

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 11

0

500

1000

1500

2000

2500

without TDRT

with TDRT

Ro

ot

me

an

sq

ua

re d

evia

tio

n

(pxl)

p < .05n.s.

p < .01 p < .001

n.s.p < .05

Secondary Task hard   hard      N-back SuRT

Primary Task easy   easy hard   easy   hard Visual Tracking   Auditory Tracking

Effect of TDRT on primary task performance: - Tracking deviation

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 12

5

6

7

8

without TDRT

with TDRT

Me

an

re

sp

on

se

tim

e (

s)

p < .01

Secondary Task SuRT, hardPrimary Task Visual Tracking, easy

Effect of TDRT on secondary task performance:- SuRT response time

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 13

Secondary Task N-back (hard)Primary Task Visual Tracking (easy)

Effect of TDRT on secondary task performance:- N-back percentage of correct answers

0

10

20

30

40

50

60

70

80

90

100

110

without TDRT

with TDRT

Pe

rce

nta

ge

of

co

rre

ct

an

sw

ers

(%

)

n.s.

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 14

Conclusions/1

• Sensitivity in triple-task scenarios:“Primary T. + Secondary T. + TDRT”

Sensitivity of TDRT to different load levels of cognitive secondary tasks could be confirmed (N-Back easy vs. hard).

TDRT seems to be not sensitive to load levels of secondary tasks which primarily demand for visual-manual resources (SuRT easy vs. hard).

TDRT is sensitive to secondary task types which vary in type of resource demands (cognitive vs. perceptual-motor)

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 15

Conclusions/2

• Sensitivity in dual-task scenarios: “Primary T.+ TDRT”

TDRT seems to be not sensitive to load levels of the surrogate driving task (bendiness of the tracking path).

TDRT is sensitive to primary task types which vary in cognitive resource demands (auditory-cognitive-motor vs. visual-motor demands of tracking task).

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 16

Conclusions/3

• Intrusiveness:

There are indications of the TDRT’s intrusiveness on primary task performance (tracking deviation).

No clear picture for intrusiveness of TDRT on secondary task performance. Intrusiveness of TDRT seems to depend on type of secondary task:

No intrusion on N-Back performance Intrusion on SuRT performance

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 17

Conclusions/4

• Further research recommended:

Interference between TDRT and secondary/primary task due to motor task demands.

Intrusiveness of TDRT on primary task and secondary task performance:

Comparative studies on degraded vs. normal task performance.

Limitations of applicability of TDRT (e.g. in on-road tests).

Sensitivity of TDRT to load levels of other secondary task types (generic tasks; real tasks)

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 18

Federal Highway Research Institute

Thank you for your attention!

European Conference on Human Centred Design for ITS, June 2014, ViennaSchindhelm/Schmidt 19

Schindhelm/Schmidt 20

Back up

European Conference on Human Centred Design for ITS, June 2014, Vienna