Implementation of a Self-Learning Route Memory for Forward-Looking Driving

In this paper it will be shown how a database containing information of the road characteristics of a frequently driven route can be automatically generated and continually updated in a vehicle during each drive. The contained information can be used as foresight information in predictive driving strategies. By using only drive train information, standard sensors (e.g. from ESC and ABS), and a GPS relevant road characteristics (curves, slopes, speed limits, etc) can be identified during the drive, stored in an on-board database, and used to optimize fuel consumption or driving comfort in subsequent trips along the route. The system is verified using a driving simulator with a 3D surround graphics system.