Using real road driving data to calibrate a model of front-end collision risk

Robert Baran, Changwon Jeon, Hanseok Ko

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Deployment of first generation collision avoidance systems offers a potential source of abundant data that can be analyzed to assess the dependence of front-end collision risk on speed and headway distance. Passive persistent monitoring of moving cars via wireless telemetry could produce large amounts of realistic data to refine collision warning and auto-braking algorithms. Our objective is to estimate an instantaneous risk function that gives the probability of hard braking conditioned on sudden deceleration to avoid collision with the vehicle ahead. We develop an abstract stochastic model in which this risk depends on two predictors, speed and time-tocollision (TTC), having the same functional form inferred from controlled tests but with parameters that may vary between drivers, roads and traffic environments. The abstract model provides a basis for re-estimating the instantaneous risk by two approaches: (1) Fit a proportional hazards model to the times between hard braking events with speed and TTC as covariates. (2) Minimize the dissimilarity of observed and expected distributions of the retrospective risk, which is instantaneous risk sampled only at hard braking times. Both approaches appear to require an iterative procedure for adaptive parameter estimation via global nonlinear function optimization. One such procedure, illustrated for the second approach with a chi-squared dissimilarity measure, typically attains 5% accuracy in parameter estimation given simulated data describing a few thousand hard braking events.

Original languageEnglish
Title of host publication6th Biennial Workshop on DSP for In-Vehicle Systems and Safety 2013, DSP 2013
PublisherKorea University
Publication statusPublished - 2013 Jan 1
Event6th Biennial Workshop on Digital Signal Processing for In-Vehicle Systems and Safety 2013, DSP 2013 - Seoul, Korea, Republic of
Duration: 2013 Sep 292013 Oct 2

Other

Other6th Biennial Workshop on Digital Signal Processing for In-Vehicle Systems and Safety 2013, DSP 2013
CountryKorea, Republic of
CitySeoul
Period13/9/2913/10/2

Fingerprint

Braking
Parameter estimation
Deceleration
Telemetering
Collision avoidance
Stochastic models
Hazards
Railroad cars
Monitoring

ASJC Scopus subject areas

  • Automotive Engineering
  • Safety, Risk, Reliability and Quality

Cite this

Baran, R., Jeon, C., & Ko, H. (2013). Using real road driving data to calibrate a model of front-end collision risk. In 6th Biennial Workshop on DSP for In-Vehicle Systems and Safety 2013, DSP 2013 Korea University.

Using real road driving data to calibrate a model of front-end collision risk. / Baran, Robert; Jeon, Changwon; Ko, Hanseok.

6th Biennial Workshop on DSP for In-Vehicle Systems and Safety 2013, DSP 2013. Korea University, 2013.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Baran, R, Jeon, C & Ko, H 2013, Using real road driving data to calibrate a model of front-end collision risk. in 6th Biennial Workshop on DSP for In-Vehicle Systems and Safety 2013, DSP 2013. Korea University, 6th Biennial Workshop on Digital Signal Processing for In-Vehicle Systems and Safety 2013, DSP 2013, Seoul, Korea, Republic of, 13/9/29.
Baran R, Jeon C, Ko H. Using real road driving data to calibrate a model of front-end collision risk. In 6th Biennial Workshop on DSP for In-Vehicle Systems and Safety 2013, DSP 2013. Korea University. 2013
Baran, Robert ; Jeon, Changwon ; Ko, Hanseok. / Using real road driving data to calibrate a model of front-end collision risk. 6th Biennial Workshop on DSP for In-Vehicle Systems and Safety 2013, DSP 2013. Korea University, 2013.
@inproceedings{75cd9604c6964197a67211431fabb250,
title = "Using real road driving data to calibrate a model of front-end collision risk",
abstract = "Deployment of first generation collision avoidance systems offers a potential source of abundant data that can be analyzed to assess the dependence of front-end collision risk on speed and headway distance. Passive persistent monitoring of moving cars via wireless telemetry could produce large amounts of realistic data to refine collision warning and auto-braking algorithms. Our objective is to estimate an instantaneous risk function that gives the probability of hard braking conditioned on sudden deceleration to avoid collision with the vehicle ahead. We develop an abstract stochastic model in which this risk depends on two predictors, speed and time-tocollision (TTC), having the same functional form inferred from controlled tests but with parameters that may vary between drivers, roads and traffic environments. The abstract model provides a basis for re-estimating the instantaneous risk by two approaches: (1) Fit a proportional hazards model to the times between hard braking events with speed and TTC as covariates. (2) Minimize the dissimilarity of observed and expected distributions of the retrospective risk, which is instantaneous risk sampled only at hard braking times. Both approaches appear to require an iterative procedure for adaptive parameter estimation via global nonlinear function optimization. One such procedure, illustrated for the second approach with a chi-squared dissimilarity measure, typically attains 5{\%} accuracy in parameter estimation given simulated data describing a few thousand hard braking events.",
keywords = "Braking, Collision avoidance, Forward collision warning, Risk assessment, Time to collision",
author = "Robert Baran and Changwon Jeon and Hanseok Ko",
year = "2013",
month = "1",
day = "1",
language = "English",
booktitle = "6th Biennial Workshop on DSP for In-Vehicle Systems and Safety 2013, DSP 2013",
publisher = "Korea University",

}

TY - GEN

T1 - Using real road driving data to calibrate a model of front-end collision risk

AU - Baran, Robert

AU - Jeon, Changwon

AU - Ko, Hanseok

PY - 2013/1/1

Y1 - 2013/1/1

N2 - Deployment of first generation collision avoidance systems offers a potential source of abundant data that can be analyzed to assess the dependence of front-end collision risk on speed and headway distance. Passive persistent monitoring of moving cars via wireless telemetry could produce large amounts of realistic data to refine collision warning and auto-braking algorithms. Our objective is to estimate an instantaneous risk function that gives the probability of hard braking conditioned on sudden deceleration to avoid collision with the vehicle ahead. We develop an abstract stochastic model in which this risk depends on two predictors, speed and time-tocollision (TTC), having the same functional form inferred from controlled tests but with parameters that may vary between drivers, roads and traffic environments. The abstract model provides a basis for re-estimating the instantaneous risk by two approaches: (1) Fit a proportional hazards model to the times between hard braking events with speed and TTC as covariates. (2) Minimize the dissimilarity of observed and expected distributions of the retrospective risk, which is instantaneous risk sampled only at hard braking times. Both approaches appear to require an iterative procedure for adaptive parameter estimation via global nonlinear function optimization. One such procedure, illustrated for the second approach with a chi-squared dissimilarity measure, typically attains 5% accuracy in parameter estimation given simulated data describing a few thousand hard braking events.

AB - Deployment of first generation collision avoidance systems offers a potential source of abundant data that can be analyzed to assess the dependence of front-end collision risk on speed and headway distance. Passive persistent monitoring of moving cars via wireless telemetry could produce large amounts of realistic data to refine collision warning and auto-braking algorithms. Our objective is to estimate an instantaneous risk function that gives the probability of hard braking conditioned on sudden deceleration to avoid collision with the vehicle ahead. We develop an abstract stochastic model in which this risk depends on two predictors, speed and time-tocollision (TTC), having the same functional form inferred from controlled tests but with parameters that may vary between drivers, roads and traffic environments. The abstract model provides a basis for re-estimating the instantaneous risk by two approaches: (1) Fit a proportional hazards model to the times between hard braking events with speed and TTC as covariates. (2) Minimize the dissimilarity of observed and expected distributions of the retrospective risk, which is instantaneous risk sampled only at hard braking times. Both approaches appear to require an iterative procedure for adaptive parameter estimation via global nonlinear function optimization. One such procedure, illustrated for the second approach with a chi-squared dissimilarity measure, typically attains 5% accuracy in parameter estimation given simulated data describing a few thousand hard braking events.

KW - Braking

KW - Collision avoidance

KW - Forward collision warning

KW - Risk assessment

KW - Time to collision

UR - http://www.scopus.com/inward/record.url?scp=84907316991&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84907316991&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84907316991

BT - 6th Biennial Workshop on DSP for In-Vehicle Systems and Safety 2013, DSP 2013

PB - Korea University

ER -