search

Press ENTER to search or ESC to close

Shape Sensor

Nova Scientia shape sensing technology measures in real time the shape deforming in a pre-defined plane or in space with high accuracy, high repeatability, high frequency and down to nano strain resolution.
youtube
Andrea Mannarino - Director of Data Science
Hello and welcome to Nova Scientia,
I’m Andrea and today I will demonstrate you our two dimensional shape sensing technology.

Setup
I will now introduce our test setup: our acquisition system; a one meter long shape sensing element; our calibration table that we will use to validate our measurements and pins that we will use to force the shape sensing element in some pre-determined shapes.
On the screen, our visualisation software where at the top you can visualise the 3D shape and at the bottom, the 2D shape. The shape sensing element is represented by the blue light and you can also visualise the position of the pins, if required.
Test
I will now demonstrate our technology: I will move the shape sensing element to the maximum positive position.
As you can see from the screen, we are able to reach the target that we set.
I will now let the shape sensing element free and we can visualise it moving on the screen in real time.
We are able to reconstruct shapes of any kind, so I can perturb the shape sensing element at any point and we are able to visualise the actual shape.
We are able to reconstruct shapes with an error of 1 mm over 1 m and with a resolution of 20 nanometres.

In this case, the shape sensing element is 1 m long, but we will be happy to customise it to any application that you have in mind.

Key Features

Key Features
• 0.1% Accuracy
Maximum error 1mm over 1m

• 20nm Resolution 
Independently from Sampling
Frequency

• High Sampling Frequency
Up to 205kHz

• Fully customisable dimensions

Shape Sensor 2D: Case Study

In this case study, a Shape Sensor 2D is used to reconstruct the shape of a 1m-long Aluminium beam clamped at one end.
The tested shapes are reconstructed accurately, with a maximum error of 1mm over the entire length of the beam.
The achieved resolution is 20nm, independently from the sampling frequency used.

Test Description

The FSI sensors are bonded onto an Aluminium beam with thickness of 2mm and length of 1000mm, for a total weight of 50g. The FSI sensors are distributed on the Aluminium beam to allow the reconstruction of the beam position every 100mm along its length, starting from its clamped end.
The assembly of the aluminium beam and the FSI sensors forms what we call shape sensor.
An internally designed calibration table is used to validate the shapes reconstructed by the shape sensor.
Pins are used to force the shape sensor position at pre-determined points, as shown in Figure 1, allowing a complete accuracy assessment.
Figure 1: Shape sensor mounted on calibration rig.Figure 1: Shape sensor mounted on calibration rig.

Test Results

The shape sensor is calibrated on Position 3 (Figure 1, bottom right). The reconstruction error is calculated as the absolute difference between the nominal Y-axis displacement at the pin location and the Y-axis displacement reconstructed by the shape sensor. From Figure 2, the reconstruction error is negligible (shown by the Pin Error variable at the top of the figure).
Figure 2: Reconstruction error of position 3Figure 2: Reconstruction error of position 3
After having calibrated the shape sensor, the other positions depicted in Figure 1 are tested. The results are shown in Figure 3. Again, the reconstruction error is small compared to the amplitude of the achieved deflections, with a maximum error of less than 1mm over the entire length of the shape sensor.
Figure 3 also shows that the shape sensor can reconstruct any type of curvature – not only monotonic curvatures like in Positions 1 to 4, but also more complex shapes like in Position 5.
Figure 3: Reconstruction error on the other positions testedFigure 3: Reconstruction error on the other positions tested
Figure 4: Measurement noise when the shape sensor is laying on a pin.Figure 4: Measurement noise when the shape sensor is laying on a pin.
Figure 4 shows the measurement noise of one of the reconstructed points when the shape sensor is laying on a pin. The oscillation of the data around its mean value is very small – extracting the standard deviation of this signal, it can be found that the resolution of our shape sensing measurements is 20nm, and it is independent from the sampling frequency used to acquire the data.