Additive manufacturing is a process that is commonly used nowadays and is at the origin of the industry 4.0. This process is mostly used in industry because of its multidisciplinary aspectand various advantages that 3D printing represents.

The process of tracking the motion of the robot is common and essential for industrial robots. Therefore, various standards and techniques have already been implemented. However, most of them deals with the general aspect of tracking the motion of the robot but few of the techniques consider the parameters and the conditions of 3D printing.

In this thesis the accuracy of a robot is studied. The measurements are done during a simulation of 3d printing process using different parameters such as the weight of 5 kg put on the end-effector of the robot, the height of 1 mm of a layer, the distance and angle of the end-effector from the base frame of the robot. To know if the robot is accurate, the robot performs a linear trajectory programmed offline on the software RobotStudio above an angle plate depending on the above parameters. For a specific choice of parameters, it is called configuration. The main goal is to measure the deviation of the online trajectory by using two axes: the y-axis and the z-axis and compared it to the offline robot path. Through the method called path comparison method, two laser distance sensors are used as a tool and are fixed on the end-effector of the robot and point at the y-axis and the other one on the z-axis. These ones display on MATLAB, graphics and raw numerical data that show the distance between the end-effector and the angle plate. The values are used to compare the online trajectory from the offline one. Also, it should be possible through various experiments to measure the influence of the above parameters on the accuracy of the robot.

In the results part, the path accuracy is studied through some graphics representing the average of 30 tests per configuration. These average graphs allow to make some conclusions about the impact of the parameters on the accuracy of the robot. Moreover, the criterion of repeatability is studied in order to check if the robot can repeat the same path. Finally, the path positioning accuracy is calculated.

From the experiments, it is possible to see that the weight and the angles of the end-effector compared to the base of the robot impact the accuracy of the robot. However, about the global accuracy of the robot, it is not possible to conclude due to a gap of value between the online and offline results. The main reason is the uncertainties caused by the set-up.Therefore, even if it is not possible to conclude about the accuracy of the robot using this method, this method can give some hints about potential errors. An error caused by the definition of the target points has been discussed. For each configuration, an adjustment of the position of the target points is to be made so that the start position coincides with the offline position. About the path positioning accuracy, it is possible to conclude that there is a path positioning accuracy of +/- 0.072 mm for the y-sensor and +/- 0.078mm for the z-sensor compared to the planned path.

Thus, for further studies, it would be interesting to study more thoroughly the reasons of the inaccuracy. It would also be interesting to use other devices such as cameras to check the accuracy of the robot in the context of additive manufacturing.