Sensor Fusion and its Importance

Sensor fusion is not a well understood concept, even among embedded designers. You hear about it often in the content of data processing, but there is still a lot of ambiguity about what it really stands for. One reason could be the confusion with the concept of data fusion.

Basically, sensory fusion is the science of combining sensory data from different sources so that the data thus derived is more dependable, more accurate and more complete, compared to using the data individually. Sensor fusion is actually a part of a bigger field called data fusion, which encompasses different ways of combining data to refine information quality or derive more information about it.

Interestingly, sensor fusion was not invented by embedded engineers or software developers. The concept has been around for a long time. In ancient times, ship captains used something called dead reckoning, which combined information from a compass and chronometer with the speed of the ship to determine their position in the sea. Sensor fusion has vast applications and is used for everything from flying airplanes to drones. Embedded developers would do well to understand and learn its applications.