Our group was given a task of creating a mathematical model, implementing a program, and correcting errors using data produced by a robot or 'puck'.  The puck is used to scan objects for imperfections.  The puck's transciever scans the object while its eight sensors relay information in the form of an x and y component. The model was supposed to use the given data and produce a mapping of where the puck moved.  More precisely, we wanted to plot where the transceiver on the puck had moved and correct errors in the plot. We created two models.  Both of these models were produced in MatLab. 


One model, pose, was based on a paper written by Bonarini, and the other model uses a recurrence relation to plot each sensor's path. We ran into problems with pose, because the data we were given contained a large amount of zeros which the program couldn't handle. We corrected this by making all the zeros equal a tiny sum. We, then, encountered more problems with other equations because of zeros showing up in the denominators.  We corrected these problems, but almost all the images produced weren't showing the results we wanted. 


The second model we created was able to map the puck's movements without running into any problems associated with pose. The resulting images, which are shown in the results page, accurately showed the path of each sensor. We, also, were able to measure the length of path of each sensor. This allowed us to see which path was shortest and therefore contained the most errors.  



This is a picture of the puck. Underneath it has eight sensors alligned in a circle pattern around the center of the puck. There is also a transciever off to the side.