Sensor fusion

Often one sensor or one measurement principle can not capture all the information needed to make a desired conclusion. The idea with sensor and data fusion is to use inputs from complementary information sources that when combined, can be used to draw conclusions that can not be obtained from any of the sources alone.

In robotics this is particularly important allowing the robot to explore a range of sensory input when perceiving and interacting with the world. For instance when grasping objects, tactile and vision combined is important to ensure proper grip and reliable manipulation.

ROBOTNOR researchers have over the last decades built up an extensive experience with a wide variety imaging sensor technologies as well as force, torque and tactile sensing solutions. Employing our physical understanding and models of the observed system and the measuring systems, we can utilize the strengths and weaknesses of different sensing technologies within an application domain,  increasing the robustness and accuracy of our solutions.