Desktop version

Home arrow Engineering arrow Tactile Display for Virtual 3D Shape Rendering

The Prototype

To test the real performance of the designed device we have build a prototype made up of three identical modules (Fig. 5.10 and 5.11). It is possible to analyse the components of a single module in Fig. 5.12, while Fig. 5.13 shows the prototype in four different configurations during the rendering process.

Prototype of first designed version

Fig. 5.10 Prototype of first designed version

Tilting system

Fig. 5.11 Tilting system

Discussion of Pros and Limits

The tactile interface described and developed fully meet the general aims of the project:

  • Desktop and portable: the dimensions of the whole system, including the support basis, are 500 x 350 x 400 mm and it can be used with a PC or a lap-top by means of a simple USB connection. Therefore, it is possible to use the device on a common desk and it is simple to move from a workstation to another.
  • Low cost: although it is too early to perform a cost analysis, it is possible to assert that the designed device is cheaper than those available. The absence of big structures and the development by means of simple banded sheets and commercial components allow us to strongly reduce the development costs of the device.
  • Modular: thanks to the independent and absolute configuration of the modules the device developed presents a high level of modularity and consequently a high customization level. Indeed, it is possible to adapt the number of control sectors by adding or removing modules according to application needs. We have realized the prototype with three modules because this is the minimum number to obtain an operating device. However, to increase the resolution of the system, it is possible to add further modules, thus increasing the number of control sectors as well as our control on the strip deformation.
  • Performance: as regards the specific characteristics in terms of the potentiality of representing a surface, the system is able to render tridimensional surfaces with curvature radii lower than 40 mm. This is an improvement compared with the devices, which are currently available.

To understand the real improvements obtained with the developed device, it is possible to compare it with the SATIN system, as shown in Table5.1.

By analysing the different characteristics it is possible to notice that the distance between the nodes is significantly lower than that of the SATIN system. In this manner, it is possible to ensure a higher resolution. As regards the number of nodes,the

Prototype of a single module

Fig. 5.12 Prototype of a single module

Prototype in different configurations

Fig. 5.13 Prototype in different configurations

previously developed solutions have a fixed number, which coincides with the maximum number of relative modules that can be used. Instead, the proposed system can work with a variable number of control points according to the users and application needs. The SATIN system does not provide the possibility of changing the nominal distance between the nodes, while the designed system has this important feature. Indeed, the designed system allows us to change the distance between the control sectors that allow us to increase or decrease the resolution of the rendered surface. But, to do this operation we need to disassemble the strip from the device and change the distance. Therefore, this adjustment requires stopping the device work to allow us to perform the required set-up. Furthermore, the impossibility to change the distance of the control sectors during the rendering prevents to place the control sectors in

Table 5.1 Comparison between the SATIN system and the first version of developed tactile device

SATIN system

SATIN system

First version of developed device

Dimensions of the whole system

2600 x 2100 x 1280mm

500 x 350 x 400 mm

Number of control points of the tactile interface

9

>3

Distance among the control points of tactile interface

90 mm

>40 mm

Possibility to regulate in real time the distance among the control points

NO

NO

Reached radius by the tactile interface

185 mm

40 mm

Modularity

NO

YES

Portable

NO

YES

Implementation cost

High

Low

specific, not-equidistant points of the trajectory. Indeed, the accuracy of the render depends by the distance between the inflection point of the trajectory and the control sector. If we are able to arrange the control sectors as close as possible to the inflection points, we will achieve a render with high accuracy. With the designed solution it is not possible to adjust in real time the relative distance between the control sectors and this limitation is the starting point for the development of the second version of the tactile interface.

Indeed, the first version of the tactile device presents other drawbacks. These are related to the transmission system elements. We placed the servomotors in charge of actuating the planar articulated arm at the basis of the module. This allowed us to avoid high inertial loads caused by the weight of these components. To transmit the motion from the servomotor to the element Arm2 a transmission system was needed and the pulleys-belt solution was the best choice in terms of dimensions. But, this system presents a stiffness that is too high for our aims. Another problem is represented by the mechanical plays. In actuality, the complexity of the transmission system and the high number of its elements make the device sensible to the small mechanical plays which are irrelevant when considered individually. On the contrary, they significantly influence the accuracy of the device in the whole kinematic chain.

Analysing the behaviour of the whole device in terms of stiffness could be considered as a good compromise between stiffness and dimensions. The designed solution works fine in standard situations but when we try to render surfaces with extremely low curvature radii, which require a strong deformation of the strip, or when the user applies high loads during the touch, the designed system lost accuracy because of the elements stiffness. To solve this problem, it would be sufficient to increase the stiffness of the elements but this will require an increase of the dimensions and in particular of the width of the module. This dimension strongly influences the minimum distance between the control sectors: by increasing the module width, there will be an increase of the minimum achievable distance, while there will be a decrease of the system resolution.

To overcome all these drawbacks, we have designed the second version of the device, which will represent an advance in all the fundamental aspects of the device.

 
Source
< Prev   CONTENTS   Source   Next >

Related topics