**4.1 Results**

212 Haptics Rendering and Applications

When the user desires finalize with the HDS's assistance, he must press the second button and indicates using a menu, if erase the line recently drawn or maintains it; if he decides maintain it, the virtual template will be erased and only the line drawn made remains (Figure 16.e). In order to design more complex sketches is necessary repeat all the precedent

To check that the users learn sketching faster when they are provided with force feedback, we carried out the following test: The image used in the experiment is shown in the Figure 17b, which is based on the original image shown in the Figure 17a. The test image was chosen due it has the basic components of more complex designs (lines, curves, circles, squares, etc).

(a) Original (b) Processed Image

We requested to 10 users to trace the test image seven times as precisely as possible in terms of position. Since this task should model free drawing, no restriction on completion time or effector velocity was imposed. The users were divided in two groups of 5 members. At the beginning of each experiment, the users had five minutes for drawing on the test image in order to become familiar with the HDS. The users of the first group were stimulated only with visual feedback throughout the experiment. The users of the second group were stimulated with visual feedback on every trial, and they received the force feedback only in

We used the performance parameter described by (6) to measure if user learning of drawing skills was indeed accelerated. The PSD parameter involves the time, velocity, distance and error (between the HEE Position and the point on the trajectory nearest to the HEE) when

0

*L vdt* 

*dt*

(6)

*t <sup>o</sup> SD <sup>t</sup>*

*P*

steps and depending of the ability of the user can draw more complex design.

**4. Experiment to validate sketching skills for fixed templates** 

Fig. 17. Test Image.

the user is drawing.

the second, fourth and sixth trials.

In the Figure 18, we present the normalized sum of PSD over the subjects of the two groups: the PSD sum for trial 1 has been made equal to 1.

Fig. 18. Evolution of normalized sumatory of the PSD

We can obviously observe a strong difference between performances on trial 2, 4 and 6. What is more interesting, we can observe that for the Visual Feedback Only group, learning stops at Trial 4, while the Visual Feedback and Haptics group continues learning to Trail 7. Furthermore, at trial 7 the normalized sum of PSD is lower for the Visual Feedback and Haptics group, which leads us to conclude (within the limits of a small scale experiment with a small number of subjects) that Haptic feedback enables learning for a longer time and leads to better results for the task at hand.

We didn't validate the improvement of the user to drawn simple skills due there aren't restrictions about position, size of the sketch's traces. It is our first attempt to develop a more advanced assistant, which will have more virtual geometric templates and the auditory cues will be exploited more properly, i.e. vocal instructions.
