An In-Vehicle Lane Departure and Erratic Driving Warning System using V2V Communication and Standard GPS Technology

Abstract

Some of the critical features of Advanced Driver Assistance Systems (ADAS), including unintentional lane departure warning and erratic driving warning have significant potential to reduce crashes. Generally, these systems use either various image processing techniques or Global Positioning System (GPS) technology with lane-level resolution maps. However, these are expensive to implement as well as have some limitations, such as harsh weather or irregular lane markings can drastically reduce their performance. Previously, we developed a lane departure warning system (LDWS) where we generated road reference heading (RRH) form a vehicle’s past travel trajectories acquired by GPS to detect unintentional lane departure. But when a vehicle travels for the first time on a given road, it does not have any past trajectory of that road to generate RRH of that road needed to detect unintentional lane departure. So, in this thesis, we have improved our previously developed LDWS by adding a vehicle to vehicle (V2V) communication feature to the existing LDWS so that a vehicle traveling on a road for the first time can acquire the RRH of that road from a nearby vehicle via V2V communication. Furthermore, we have also enhanced the existing LDWS by adding a parallel erratic driving warning system (EWDS) to detect erratic driving behavior of a vehicle so that the system can issue timely warnings to alert the driver. We have considered two most common erratic driving scenarios; inter lane change and intra lane change erratic driving. We have developed an algorithm to detect both erratic driving behaviors and implemented the algorithm in a prototype system. We have extensively tested the V2V communication feature of LDWS as well as theEDWS in the field to evaluate their accuracy in real-time. Our field test results show an RRH successfully transfers from one vehicle to another, and the EDWS can detect each erratic driving scenario during the test drives accurately in a timely manner.