Evaluation of an App-Based Intervention Program for Teen Driver Speeding Mitigation and Intelligent Speed Assist Technology Use: Study Findings

2026 Discussions Advancing Research in Transportation Safety

Saturday, January 10, 2026

Renée St. Louis (PI),

Jennifer S. Zakrajsek,

Nicole Zanier, David W. Eby, Boluwatife A. Dogari

University of Michigan Transportation Research Institute

John K. Lenneman Ananna Ahmed

Toyota Motor North America Collaborative Safety Research Center

Background / Problem Statement

3

Speeding was a factor in 29% (11,775) of total fatalities in 2023.

Source: NSC Injury Facts 2025.

Source: NSC Injury Facts 2025.

Teen Driving Behavior

• Teens more likely to speed and tailgate

4

• Fatal crash rate for 16-19 years old ~3x higher than 20 years old
• Positive benefits for parental involvement

Percentages of Speeding Drivers Involved in Fatal Traffic Crashes, by Age Group and Sex, 2022

Intelligent Speed Assist (ISA) Systems

• Uses mapped speed limit data or information extracted from road signs to ensure awareness of speed limit and facilitate speed management

5

• ISA escalation potential
• Inform
• Warn
• Actively prevent speeding

Project Aims

6

• Evaluate effect of ISA parental involvement intervention on teen ISA use (in-vehicle or app-based)
• Evaluate teen driving behavior in relation to intervention implementation and teen ISA use

The Intervention

7

Developed and implemented an intervention that:

• Uses a parent-teen driving agreement
• Facilitates parental monitoring of teens’ driving performance scores from a smartphone app
• Coaches parents to give feedback about teens’ driving scores to encourage teens to drive safely

Parent-Teen Driving Agreement – PULSE Pact

8

Monitor Driving Performance – OtoZen App

9

• Uses sensor data (accelerometer, GPS, mapping databases) to calculate a risky driving score for each trip (ranging from 0-100)
• Scores calculated for each trip based on:
• Frequency, Magnitude, and Duration of speeding events
• Cell phone use while driving 10+ mph
• Location of speeding events and cell phone usage recorded and provided as feedback
• Logs of driving trips exported
• Parents and teens given instructions on how to download and set up the app

10

Feedback – Parent Guidance Document

11

• Provides guidance for using OtoZen and giving feedback to teens
• Rationale for document design
• Pilot study
• Theory of Planned Behavior

​

12

Example feedback from Parent Guidance document:

Research shows that you will be most likely to change your teen’s unsafe driving behaviors by trying to change:

13

Methods

14

Participants

15

• Dyads resided throughout Michigan
• All teens were at GDL Level 2 or Level 3 (independent license)

​

Procedure

16

Quantitative Analysis Measures

17

Results

18

The Sample

19

• Study Enrollment
• 30 parent-teen dyads: 12 In-Vehicle ISA, 18 OtoZen App ISA
• One dyad was lost, leaving 29 for the final analysis (12, 17)
• Parent Demographics
• Mean age = 47.3 years (range: 39-62)
• Mostly women, highly educated (bachelor’s or higher), annual household income >$80,000
• Teen Demographics
• Mean age = 16.4 years (range: 16-17)
• Balanced across gender

​

Driving History and Habits (Pre-Study)

20

• Parents’ Self-Reported Driving
• Drove 6.5 days and 207.4 miles per week
• 30.9 years of independent driving experience
• Teens’ Self-Reported Driving
• Drove 5.9 days and 91.3 miles per week
• 1.2 years of independent driving experience
• Road Type Usage
• More of teens’ driving was on local streets and arterials than parents’

​

Teens’ Observed Driving During the Study

21

Intervention Engagement by Teens

22

Both groups had ISA on and used alerts. Use was higher and more consistent for the In-Vehicle ISA group

Intervention Engagement by Parents

23

App engagement high for both groups. Feedback more balanced and consistent for In-Vehicle ISA group.

How ISA Helped Teens

24

Both groups reported multiple benefits; more In-Vehicle users reported benefits, but App users’ benefits increased throughout intervention

How ISA Impacted Teens

25

Decreased speed was most frequent for both groups; App users reported fewer negative effects

TPB Constructs: Speeding and Tailgating

26

• Teens
• Rated speeding more dangerous (p=.06)
• Rated speeding less pleasant (p=.02)
• Parents
• Rated speeding more dangerous (p=.01)
• Rated speeding less pleasant (p<.01)

​

• Teens rated tailgating more dangerous (p=.04)

​

Teens and parents viewed speeding less favorably after intervention

TPB Constructs: Compared by Group

27

• Following the intervention, teens and parents in both ISA groups viewed speeding as more unpleasant
• This change was greater for teens in the In-Vehicle ISA group (H=6.2, p=.01)
• This change was greater for parents in the App ISA group (H=4.4, p=.03)
• At post: Statistically significant differences between groups: 75% of In-Vehicle ISA teens reported Rarely speeding while 29% of App ISA teens reported Rarely speeding (χ²=10.6, p=.01)

​

While most measures did not change differently by ISA group, there were some notable differences

Observed Driving Measures by ISA Group

28

• Teens’ trip scores started high at baseline and remained high during the intervention period
• The intervention did not significantly change teens’ trip scores or speeding events
• Phone use events per trip significantly decreased for all teens (-0.09, p=.01), an effect driven by the In-Vehicle ISA group (-0.14, p=.004)

​

Comparison of baseline and intervention

trip scores by ISA group

Comparison of baseline and intervention

phone use events by ISA group

Correlation Analysis

29

• Spearman Rank-Order correlation analysis performed
• Examined the relationship between driving outcomes (speeding, phone use, trip scores) and various predictor variables
• Predictors included:
• Trip characteristics (from app data)
• Intervention engagement (from brief surveys)
• Demographics, attitudes, and psychosocial factors (from baseline and post surveys)
• Prior adverse driving events (from driving records)
• Teen’s study group

​

Significant Correlations with Observed Driving

30

• Parent app engagement significantly correlated with intervention trip score
• Parent feedback about app scores: 0.45 (p=.01)
• Parent viewed app: 0.32 (p=.08)
• Parent feedback correlated with intervention speeding events
• Feedback – perceived behavioral control: -0.32 (p=.09)
• Teen post-study perceived behavioral control significantly correlated with intervention speeding events: -0.49 (p=.01)

​

Parent engagement and teen perceived behavioral control correlated with better scores and fewer speeding events

Benefit of Parental Involvement

31

Qualitative Analysis Highlights

32

• Teens reported conversations likely would not have happened without the PTDA, and to stay positive with discussion
• OtoZen App ISA teens suggested integrating the app features directly into vehicles (similar to in-vehicle systems)
• Some parents reported that knowing their teen used ISA reduced their stress

​

Key Takeaways

33

• Remote administration of the study was feasible
• ISA associated with some improved safety behaviors and attitudes
• Parental involvement is a factor in those improvements
• The combination of ISA technology and active parental involvement can yield greatest benefits
• App-based ISA can be effective when in-vehicle ISA is not an option
• The program may offer non-driving psychosocial benefits by increasing positive parent-teen interaction

​

Challenges, Limitations, and Future Needs

34

• Small sample size
• Limited representiveness of sample
• High starting level of teens’ safe driving
• Limited follow-up period
• Future research:
• Better understanding of varied access to ISA and how best to leverage in-vehicle and smartphone-based systems
• Explore how to optimize parental influence
• Examine long term impact and how to sustain potential benefits

​

35

Thank you!��Questions?��