24 IEEE Instrumentation & Measurement Magazine June 2008 1094-6969/08/$25.00©2008IEEE 14 Fourteenth in a Series Sensor Input Processing Output Display, actuators, signals, control M ultisensor data fusion is a multilevel, multifaceted process dealing with the automatic detection, association, correla- tion, estimation, and combination of data from single and multiple information sources. The results of a data fusion process help users make decisions in complicated scenarios. Integration of multiple sensor data was originally needed for military applica- tions in ocean surveillance, air-to- air and surface-to-air defense, or battlefield intelligence. More recently, multisen- sor data fusion has also included the nonmilitary fields of remote environ- mental sensing, medi- cal diagnosis, automated monitoring of equipment, robotics, and automotive systems. Inputs to a multisensor data fusion system include raw sensor data, commands, model parameters, and other a priori information. Output is much more heterogeneous and can range from simple estimates about the attributes of a certain entity (a physical phenomenon, a military target, or a fault condition in a machine) to complex inferences about current or future relationships between multiple entities. During the past 15 years, several data fusion methods have been proposed. This article describes some data fusion models and some applications to next-generation car safety and driver assistance systems. These applications are par- ticularly suitable to provide an overview of multisensor data fusion starting from the plain detection of multiple objects around a given host vehicle to inferring (in order of increasing complexity): The position and speed of possible obstacles The type of the objects in the road environment The relative movement and distribution of ob- stacles over a given area The early detection of a possible collision Possible suggestions for prompt and effective counter- measures (e.g., sudden braking, steering wheel adjustment, etc.). For instance, the techniques developed within the European Union project titled “Preventive and Active Safety Applications Contribute to the Road Safety Goals on European Roads” (PReVENT) are able to create safety zones around vehicles by means of embedded multisensor data fusion systems that sense the type and ◗ ◗ ◗ ◗ ◗ Tutorial 14: Multisensor Data Fusion Part 14 in a series of tutorials in instrumentation and measurement David Macii, Andrea Boni, Mariolino De Cecco, and Dario Petri