A COMPARISON OF VEHICLE ALERT MODALITIES' TIME-TO-COLLISION WARNINGS TRIGGERED BY THE VEHICLE'S CONTROLLER AREA NETWORK SYSTEM

摘要

Currently, the time-to-collision (TTC) is determined as the time when external instrumentation measures a data flag from the Controller Area Network (CAN) signal or at the time an alert modality can be used to evaluate the performance of a vehicle's Forward Collision Warning (FCW) system for the National Highway Traffic Safety Administration's (NHTSA's) New Car Assessment Program (NCAP). Many vehicle manufacturers assess FCW performance using the digital signal from the CAN to determine the onset of a warning which can then be used to determine compliance with TTC timing requirements listed in NHTSA's performance test procedure provided at www.regulations.gov in docket number NHTSA- 2006-26555-0128. NHTSA has observed that the onset of an FCW alert can be substantially delayed when compared to the activation time of the CAN signal. The purpose of this paper is to compare the timing of the CAN signal to the actual visual and audial alerts obtained during the same trial, to determine the extent of these differences, and how they vary by vehicle manufactuer. The CAN signal and two alert modalities (visual and sound) for seven vehicles were collected by Dynamic Research, Inc., and the subsequent TTCs were calculated using the test procedures and equations established by the agency. Data from the seven vehicles were analyzed for three separate test configurations. Initial analysis did not separate the vehicles by manufacturer; however, upon noticing a linear trend between the CAN signal and visual alerts, the data was grouped by manufacturer for further analysis. A strong linear relationship (R~2>0.8) was discovered between visual and CAN signal warnings, which correlates to a constant amount of delay between the CAN and visual alerts for all seven (7) vehicles as well as the audial and CAN signal warnings for four (4) of the seven (7) test vehicles. For the remainder of the vehicles, an insonsistent delay was exhibited within models. The audial-CAN relationship was not discovered until vehicle data was separated by manufacturer. Vehicles that exhibited a constant delay from when the CAN data flag was issued to when the visual or the audial alert was measured were more likely to pass the TTC requirements. Certain models had visual and audial alert modalities occur after the minimum safe TTC has passed. As a result, this paper will also attempt to conjecture potential reasons for the differences delay in the FCW alert modalities timing compared to that of the CAN data flag.