The Intelligent Systems Research Lab (ISRL) at the Electrical Engineering and Computer Engineering department at the University of Michigan – Dearborn (UMD) has been actively performed in the research areas of machine learning and computer vision with applications to automotive industry. The ISRL is directed by Professor Yi Lu Murphey and currently has two research scientists, three PhD students, three Master's students and one undergraduate student. During the past 20 years, the ISRL has performed over 40 projects funded by government agencies (NSF, DoD, State of Michigan) and industrial companies (Ford Motor Company, TRW Corporation, Johnson Control, Jabil Circuit, etc), collaborated with other researchers and produced many peer-reviewed publications.
Please visit our website for more details.
Accurate prediction of traffic information is important in traffic control systems, optimizing vehicle operations, and the individual driver. The predicted traffic information can reduce the uncertainty of the future traffic states and provides the driver with useful information such as expected delay and alternative routes to the destinations. But the prediction the traffic information is a challenging problem due to many dynamic contributing factors in traffic.
In 2010, ISRL has started the research of the Intelligent Dynamic Trip Modeling (IDTM) system for vehicle speed profile generation. This research is supported in part by a research contract from Ford Motor Company. IDTN is to predict the speed profile for the given route when the driver entered the origin and destination in the navigation system. To build the IDTM system successfully with high accuracy, reliable traffic data source and on-site experiments are essential parts. NAVTEQ ADASRP provides facilities that represent source data and on-site observation. IDTM is planning to use the NAVTEQ traffic data source to train the IDTM system and generate the dynamic speed profile prediction using the current traffic information. IDTM will be developed based on NAVTEQ ADASRP.
