This project is in cooperation with Prof. R. Wehner (this department) and with the Artificial Intelligence Lab (Department of Computer Science at the University of Zurich) www.ifi.uzh.ch/groups/ailab/

A wheeled, mobile robot was equipped with three polarization-opponent sensors (POL-OP units; see Fig 3 ) similar to the model POL-neuron presented above but directed to the zenith. The units were tuned to e-vector orientations parallel, 60° and 120° to the robots' long axis ( Fig. 4 ). Using the output signals of the POL-OP units ( Fig. 5 ), there are two ways by which the robot can determine its body orientation with respect to the e-vector orientation in the zenith. (1) Sequential method: The robot rotates about its vertical axis until one of the POL-OP units shows maximal response. (2) Instantaneous method: The robot calculates its orientation by comparing the signals of the three differently tuned POL-OP units. It is still unknown which one of the two methods the insects employ.

We have built two versions of robot: Whereas in Sahabot 1 the polarization compass just served to keep a defined direction of motion, Sahabot 2 ( Fig. 6 ) was equipped with a path integration mechanism that constantly informed the robot about its momentary position and orientation. (Sahabot 2 also contained a camera system allowing landmark navigation). The robot was given certain navigational tasks and its performance was assessed. We found that the robot performed well with both the scanning and the simultaneous method of signal evaluation, although orientation was slightly more exact with the latter mechanism. The navigational precision was comparable to that of the desert ant Cataglyphis bicolor. We also found that precision of orientation dropped dramatically when the polarization compass was replaced by proprioreception (via the robots' wheels).

_{Lambrinos D, Maris M, Kobayashi H, Labhart T, Pfeifer R, Wehner R (1997) An autonomous agent navigating with a polarized light compass. Adaptive Behavior 6:131-161}

_{Lambrinos D, Möller R, Labhart T, Pfeifer R, Wehner R (2000) A mobile robot employing insect strategies for navigation. Robotics Autonom Syst 30:39-64}