A Haptic Feedback Actuator Suitable for the Soft Wearable Device

Round 1

Reviewer 1 Report

This paper describes the design, manufacturing, and testing of a novel electro-hydraulic actuator for haptic feedback, that can potentially overcome some limitations of current technologies.

The paper is well-written and of interest for the scientific community. However, I believe some issues have to be addressed before the paper can be considered for publication:

• what is the purpose of showing equations 1 and 2? I believe the authors should also prove that their device follows these equations and use these equations to tailor the device performance
• Figure 2: I suggest you add a,b, c,....also H and deltaH are not clearly visible in the schematic representation
• Figure 8: by "Load" you just mean an external load you applied on the top of your device? what is purpose of applying this load in terms of applications?
• Figure 10: the resolution and clarity of this figure should be improved. Also, the main goal of the test showed in Figure 10 is not clear. Do you just use the tactile sensor to trigger the actuation of your haptic feedback actuator? If so, this is not enough. You should prove that the output displacement/force of your device is proportional to the pressure/displacement applied during grasping and tailor your performance using theory+control laws, otherwise this test is pointless.

Author Response

Response to Reviewers’ Comments

General Response:

Author Reply: We would like to thank you for giving us a chance to resubmit the paper, and also thank the reviewers for giving us constructive suggestions. Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made thorough corrections which we hope meet with approval. We mark all the changes highlighted in the file manuscript.R1_Highlighted. All the issues raised by the reviewers have been addressed item by item.

 Item-by-item Response:

Reviewer #1:

This paper describes the design, manufacturing, and testing of a novel electro-hydraulic actuator for haptic feedback, that can potentially overcome some limitations of current technologies.

The paper is well-written and of interest for the scientific community. However, I believe some issues have to be addressed before the paper can be considered for publication:

 1. what is the purpose of showing equations 1 and 2? I believe the authors should also prove that their device follows these equations and use these equations to tailor the device performance.

Author Reply: The driving force of the actuator comes from the Maxwell force between the electrodes. Equation 1 describes the relevant variables that affect Maxwell force such as the electrode area, the distance between the electrodes, and the dielectric constant between the electrodes. According to the guidance of formula (1), we have determined the design principles of electrode, dielectric material and the parameters of the actuator. Formula (2) describes the theoretical calculation of the thickness variation after applying a voltage. The specific size parameters of the actuator are determined by the formula. The relevant characteristics of the actuator can be analyzed in an ideal state. Therefore, the formula could qualitatively analyze the relevant parameters for the output performance of the actuator.

2. Figure 2: I suggest you add a,b, c,....also H and deltaH are not clearly visible in the schematic representation.

Author Reply: We have labelled a, b, c and remarked the parameters of the actuator in Figure 2. H and DH are now clearly visible in the schematic representation.

Figure 2. The action principle of the electro-hydraulic haptic feedback actuator.

The following text has been added into the revised manuscript:

  “where H is the arched height, D is the inner diameter of the annular electrode, and D₀ is the diameter of the inner chamber, as shown in Figure 2.”(Line 117, page 6)

3. Figure 8: by "Load" you just mean an external load you applied on the top of your device? what is purpose of applying this load in terms of applications?

Author Reply: The actuator is designed to be applied to the human hand, and the actuator will output pressure to the fingers. In order to make the experiment accord with the actual use of the actuator, “load” is placed in the center area of the actuator to simulate the contact by a human hand as shown in the following figure. “Load” is a series of blocks with different weights made by Fused Deposition Modelling (FDM). The block is placed respectively in the center area of the actuator to simulate different contact pressure.

The following text has been added into the revised manuscript:

  “Tactile experience is the coordinated response of many different types of nervous systems to pressure, temperature, pain, joint position, muscle perception and movement. The generation of tactile sensation requires the action of multiple receptors located in the skin. In the paper, the Merkel disc receptors are just stimulated, which are responsible for continuous touch and pressure perception that needs to be activated by stress. Activating the human finger's sensory receptors for touch requires the corresponding pressure stimulation. In reality, the tactile feedback actuator needs to output pressure on the finger to activate the finger’s sensory threshold to reproduce the touch. When haptic feedback actuators are used in real situations, force and displacement are often required at the same time. In order to make the experiment accord with the actual use of the actuator, a load is placed in the center area of the actuator to simulate the contact with a human hand. Loads with different weights are made by Fused Deposition Modelling (FDM) and applied to simulate different contact pressure.” (Line 229, page 13)

 Figure 10: the resolution and clarity of this figure should be improved. Also, the main goal of the test showed in Figure 10 is not clear. Do you just use the tactile sensor to trigger the actuation of your haptic feedback actuator? If so, this is not enough. You should prove that the output displacement/force of your device is proportional to the pressure/displacement applied during grasping and tailor your performance using theory+control laws, otherwise this test is pointless.

Author Reply: We have reworked Figure 10 and improved the resolution and clarity of this figure. The main purpose of Figure 10 is to prove that the actuator can output different displacements according to different gripping forces. Based on the previous research, we have fitted the relationship between the applied voltage (u) and the output displacement (X) of the actuator.

 (a= -0.00228, b=0.13171, c=0.15911)

F_max is assumed to be the maximum pressure that the robotic hand can feel when it grasps an object, and F is the force obtained by the sensor in real time. The maximum displacement of the actuator is 1.1 mm. We set the force obtained by the sensor to be proportional to the output displacement of the actuator. When the robotic hand touches an object, the output displacement of the actuator X can also be calculated by the following formula.

The relationship between the force F and the resistance R of the sensor is fitted into the following formula.

So the functional relationship between the resistance of the sensor and the applied voltage of the actuator can be determined and applied in the programs of the control system.

The following text has been added into the revised manuscript:

“During the grasping process, the output force of the actuator should be associated with the value obtained by the tactile sensor. Based on the previous research, we have fitted the relationship between the applied voltage and the output displacement of the actuator in Figure 10(b). The maximum displacement of the actuator is 1.1 mm. We set the force obtained by the sensor to be proportional to the output displacement of the actuator. According to the relationship between the force and the resistance of the sensor, the functional relationship between the resistance of the sensor and the applied voltage of the actuator can be determined and applied in the programs of the control system.” (Line 272 page 16)

  Author Response File: Author Response.pdf

Reviewer 2 Report

The authors presented an interesting actuator with a sorrow analysis of the actuator itself and a design-proposal for a complete manipulation system based on the actuator. The whole approach is relevant for researchers in the field and gives valuable information. There are two areas where an improvement would be highly recommended, but even without those improvements publication is possible.

• At two points the authors state that "20 Hz matches the haptic perception of humans". This statement is not true. 20 Hz lies within the range of haptic perception of humans, but the actual perception goes from almost as steady as 0.1 Hz (although very week) up to highest sensitivities at around 200 to 300 Hz. So the authors can not claim that they designed for a haptic perception, just by "conincidence" their actuator has its resonance frequency at 20 Hz. If they could be more precise at this point about cause and effect the authors will not expose themselves to low credibility among the haptic researchers out there. Especially the statement in 186 to 187 made me smile. If a haptic sensation would be generated by "continous pressure" we would not be able to wear clothes as our nervous system would not be able to filter these low-frequency static load. Definitely not: Haptics happens at high frequencies. Noone thinking about it would ever think of haptic to happen at low frequencies.
• The actuator discussion between page 6 and 8 contains relevant information, but it also shows that the authors need to dive a little deeper in how an actuator should be characterized if used for a direct skin contact. The authors did a no-load and load-study at static control signal and at varying control isgnals (AC). As a matter of fact, the design seems to have a tuneable stiffness (so it behaves like a spring) when operated with almost no friction and a little bit of mass included. This can be seen from the square-wave reaction diagrams they measured. If now loaded with additional mass, the system reduces amplitude. The load with mass is a nice test for an actuator, the behaviour of a human-skin is much more like a spring/damper model. Instead of the experiments shown it would have been nice to derive an equivalent spring model of the actuator, and couple it to a spring representing the stiffness of the finger. This would have given a clear idea of how displacement and force/pressure really interact with each others. The authors obviously miss and understanding how to describe a coupled dynamic actuator/load system leading to some statements which are a little strange for a haptic-researcher, like the lines 200 to 206. Haptics is not about force or displacement output. Haptics is a coupled mechanical system. You can not distinguish between in- and outputs, you can only quantify the dynamic interaction of two mechanical systems by considering relations between force and displacement. 

Author Response

Response to Reviewers’ Comments

General Response:

Author Reply: We would like to thank you for giving us a chance to resubmit the paper, and also thank the reviewers for giving us constructive suggestions. Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made thorough corrections which we hope meet with approval. We mark all the changes highlighted in the file manuscript.R1_Highlighted. All the issues raised by the reviewers have been addressed item by item.

 Item-by-item Response:

Reviewer #2:

The authors presented an interesting actuator with a sorrow analysis of the actuator itself and a design-proposal for a complete manipulation system based on the actuator. The whole approach is relevant for researchers in the field and gives valuable information. There are two areas where an improvement would be highly recommended, but even without those improvements publication is possible.

1. At two points the authors state that "20 Hz matches the haptic perception of humans". This statement is not true. 20 Hz lies within the range of haptic perception of humans, but the actual perception goes from almost as steady as 0.1 Hz (although very week) up to highest sensitivities at around 200 to 300 Hz. So the authors can not claim that they designed for a haptic perception, just by "conincidence" their actuator has its resonance frequency at 20 Hz. If they could be more precise at this point about cause and effect the authors will not expose themselves to low credibility among the haptic researchers out there. Especially the statement in 186 to 187 made me smile. If a haptic sensation would be generated by "continous pressure" we would not be able to wear clothes as our nervous system would not be able to filter these low-frequency static load. Definitely not: Haptics happens at high frequencies. Noone thinking about it would ever think of haptic to happen at low frequencies.

Author Reply: Thank you for your valuable reviews! We are sorry about the misleading expressions in the paper. (1) Indeed, the frequency of the input voltage (20 Hz) doesn’t match the haptic perception of humans. In our research, we focus on the fact that different voltage loading methods will make the actuator to show different haptic effects. The output displacement of the actuator will not remain stable by applying DC voltages, but the output displacement of the actuator can remain stable by applying AC voltages with different frequencies. The frequency (20 Hz) was chosen because it can make the actuator output enough stable displacement. (2) Tactile experience is the coordinated response of many different types of nervous systems to pressure, temperature, pain, joint position, muscle perception and movement. The generation of tactile sensation requires the action of multiple receptors located in the skin. We just stimulate the Merkel disc receptors responsible for continuous touch and pressure perception which needs to be activated by stress. The proposed actuator can output relatively stable displacement with different frequencies (1-100 Hz) as shown in Figure 5.

Some sentences have been modified to make the paper more rigorous in the revised manuscript:

“By adjusting the voltages and frequencies, a maximum output displacement of 1.1 mm and an output force of 1 N/cm² can be rapidly achieved at a voltage of 12 kV (20 Hz) to meet the human hand’s tactile perception.” (Line 26, page 2)

“The decrease in displacement with the DC voltage allows the user to obtain only a short feedback effect, while the AC voltage loading method can stably and continuously output displacement for the finger with different frequencies.” (Line 213, page 12)

“Tactile experience is the coordinated response of many different types of nervous systems to pressure, temperature, pain, joint position, muscle perception and movement. The generation of tactile sensation requires the action of multiple receptors located in the skin. In the paper, the Merkel disc receptors are just stimulated, which are responsible for continuous touch and pressure perception that needs to be activated by stress. Activating the human finger's sensory receptors for touch requires the corresponding pressure stimulation.” (Line 229, page 13)

     “By adjusting the voltages and frequencies, a maximum output displacement of 1.1 mm and an output force of 1 N/cm² can be rapidly achieved at a voltage of 12 kV (20 Hz) to meet the human hand’s tactile perception. ”(Line 321, page 19)

2. The actuator discussion between page 6 and 8 contains relevant information, but it also shows that the authors need to dive a little deeper in how an actuator should be characterized if used for a direct skin contact. The authors did a no-load and load-study at static control signal and at varying control signals (AC). As a matter of fact, the design seems to have a tuneable stiffness (so it behaves like a spring) when operated with almost no friction and a little bit of mass included. This can be seen from the square-wave reaction diagrams they measured. If now loaded with additional mass, the system reduces amplitude. The load with mass is a nice test for an actuator, the behaviour of a human-skin is much more like a spring/damper model. Instead of the experiments shown it would have been nice to derive an equivalent spring model of the actuator, and couple it to a spring representing the stiffness of the finger. This would have given a clear idea of how displacement and force/pressure really interact with each others. The authors obviously miss and understanding how to describe a coupled dynamic actuator/load system leading to some statements which are a little strange for a haptic-researcher, like the lines 200 to 206. Haptics is not about force or displacement output. Haptics is a coupled mechanical system. You can not distinguish between in- and outputs, you can only quantify the dynamic interaction of two mechanical systems by considering relations between force and displacement.

Author Reply: Thank you for your constructive suggestions! We focus on the design, the preparation, the output characteristics and the application of the proposed actuator in the paper, but we ignore the relations between force and displacement for a direct skin contact. After the finger touches the actuator, the coupled motion between the actuator and the finger is a relatively complex model for us. At present, we only use the load to simulate the pressure on the actuator when it comes in contact with a human finger. In the future, we will enhance a more in-depth understanding of the mechanism of tactile generation to improve our research, and strive to achieve more realistic tactile feedback.

The following text has been revised in the manuscript:

“The human tactile sensation comes from the external stimulation of human skin receptors. Different types of human skin receptors have different sensory thresholds. Tactile experience is the coordinated response of many different types of nervous systems to pressure, temperature, pain, joint position, muscle perception and movement. The generation of tactile sensation requires the action of multiple receptors located in the skin. In the paper, the Merkel disc receptors are just stimulated, which are responsible for continuous touch and pressure perception that needs to be activated by stress. Activating the human finger's sensory receptors for touch requires the corresponding pressure stimulation. In reality, the tactile feedback actuator needs to output pressure on the finger to activate the finger’s sensory threshold to reproduce the touch. When haptic feedback actuators are used in real situations, force and displacement are often required at the same time. In order to make the experiment accord with the actual use of the actuator, a load is placed in the center area of the actuator to simulate the contact with a human hand. Loads with different weights are made by Fused Deposition Modelling (FDM) and applied to simulate different contact pressure. Figure 8 plots the displacement as a function of load from 1 to 7 g. The output pressure of the actuator starts from 0 to a maximum of 1 N/cm2. The displacement decreases with increasing load. The maximum measured displacement of 0.15 mm is obtained from a load of 7 g.”(Line 227, page 13)

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors addressed all my concerns. 

I just have one last suggestion. While describing Figure 10, authors said "based on previous research". I believe a ref and more details (like the details provided in the authors response document" are needed.