#Staff Column
2022.07.29
Bringing to life a sensor that mimics human skin
Columnist | Introducing the columns written by member of Amorepacific Group
Sensor and cosmetics Part 2. Bringing to life a sensor that mimics human skin
Columnist | Seo Jeongeun
Amorepacific R&D Strategy Lab
Amorepacific R&D Strategy Lab
# INTRO
When we try out a new cosmetics product for the very first time, we often take a little bit of the product with a finger to gently rub and pat on the back of our hand. We remember that feeling of application as we try out different products before finding the one we like. This process of trying different products is one of the joys that come from experiencing a cosmetics product.
# The process of developing a sensor that mimics the sensitivity of human skin.
Human skin is an ideal tactile sensor. It collects information about the object being touched, including its temperature (thermal conductivity), force distribution, slipperiness, and vibration. If machines were equipped with tactile sensors that functioned in the same manner as human skin, the interaction between them and their users would be much easier.
Studies on tactile sensors involve improving the sensitivity of sensors by combining them with functional materials and novel structures, as well as bringing to life a multi-functional tactile sensor system that is able to detect a variety of stimulations, including pressure, temperature, vibration, and humidity. Sensors that mimic human touch continue to be developed in the case of humanoid robots as well.
Significant progress has been made in terms of research on sensors that are capable of distinguishing the different textures of materials, such as solids and textiles, as well as on the friction force of lubricants and liquids. However, there has yet to be a tactile sensor that detects feelings of application that vary according to the different combinations of fluid liquids in air-exposed environments. This is why we decided to develop our own sensory technology in order to digitalize feelings of application, which is critical in customer experiences.
When we try out a new cosmetics product for the very first time, we often take a little bit of the product with a finger to gently rub and pat on the back of our hand. We remember that feeling of application as we try out different products before finding the one we like. This process of trying different products is one of the joys that come from experiencing a cosmetics product.
# The process of developing a sensor that mimics the sensitivity of human skin.
Human skin is an ideal tactile sensor. It collects information about the object being touched, including its temperature (thermal conductivity), force distribution, slipperiness, and vibration. If machines were equipped with tactile sensors that functioned in the same manner as human skin, the interaction between them and their users would be much easier.
Studies on tactile sensors involve improving the sensitivity of sensors by combining them with functional materials and novel structures, as well as bringing to life a multi-functional tactile sensor system that is able to detect a variety of stimulations, including pressure, temperature, vibration, and humidity. Sensors that mimic human touch continue to be developed in the case of humanoid robots as well.
Significant progress has been made in terms of research on sensors that are capable of distinguishing the different textures of materials, such as solids and textiles, as well as on the friction force of lubricants and liquids. However, there has yet to be a tactile sensor that detects feelings of application that vary according to the different combinations of fluid liquids in air-exposed environments. This is why we decided to develop our own sensory technology in order to digitalize feelings of application, which is critical in customer experiences.
＜Figure 1. Digital transformation of manufacturing＞
The tactile sensor (Fig 2) developed by a team from the Korea Research Institute of Standards and Science (KRISS) collects a wide range of tactile data from objects, including roughness, friction, temperature, and hardness, in a similar manner as the human hand does. Because the high-functionality semiconductor silicon-based sensor stores digitized values of measurements, it is able to recognize objects even more accurately than the human hand can. According to the results of repetitive tests using approximately 25 samples, from textile to wood to plastic, the sensor yielded an over 98% accuracy. Upon hearing about this result, we scheduled a meeting with the research team. However, the sensor, although superior to the human hand in terms of detecting curves and embossing of dry surfaces, had a difficult time distinguishing among different cosmetics products. The research team responded that achieving such high sensitivity would be challenging unless there was a specialized sensor designed just for cosmetics.
＜Figure 2. A humanoid tactile robot developed by KRISS. A system that can measure surface hardness, temperature, roughness, and friction＞
After meeting with several experts, we came to realize that this was not going to be an easy task. Nonetheless, giving up was not an option. We continued to look for a partner. We then came across some outstanding technology developed by a professor from the Ulsan National Institute of Science and Technology (UNIST). We immediately requested a test to see if his sensor would be able to distinguish between three samples with different feelings to the skin upon application. The result? Not only was the test successful, but the professor had an impressive understanding of the project. We gladly waited a year for him to return from his sabbatical to embark on a co-research project.
First stop: a sensor to datafy ‘coolness’
＜Figure 3. Diagram of electronic skin that mimics the skin and its temperature sensitive channels＞
‘Coolness’ in relation to the texture of the product is defined as the feeling on the skin you get when the skin temperature drops as the water (ethanol) solvent evaporates. This feeling is among the 17 criteria of evaluation by a professional panel. (Coolness may be associated with substances like menthol, which is excluded for the purpose of this study.)
The human skin is designed to reduce the body temperature through heat vaporization, which takes away the surface heat as sweat evaporates. The market has seen an array of products that use this reaction to prevent thermal aging by delivering a feeling of refreshment and coolness to the skin. This drop in the temperature can be measured using an infrared thermal camera. Data like the contour of the area temperature may be digitized using infrared rays, but more recently, there has been active research on sensors that can both detect and datafy change in temperature in real-time using electronic skin. A highly sensitive temperature sensor is a must in measuring this coolness because the sensor should be able to detect the slight heat of vaporization that varies on the type of solvent.
The extent and retention of coolness we feel when we apply a product vary according to the frame formulation of the product. Delivering the best sensory experience and uniqueness of the product requires a comprehensive analysis of each formulation as well as of the nature of ingredients and their combinations. We sought to create a temperature change sensor that would be optimized for cosmetics analysis in order to use the accumulated data in researching and predicting new formulations.
A sensor that detects in real-time the slightest change in temperature, just like human skin
There are requirements for measuring the coolness that we feel when applying a product.
The change in temperature should be measured with less than 0.1℃ sensitivity through real-time monitoring.
There should not be any change in resistance that comes from external pressure, rubbing and or solvent (current flow).
Most conductive sensors claim that they measure heat and pressure at the same time because resistance changes according to heat, but this is not ideal. When change takes place in resistance, distinguishing the cause of the change, i.e., whether it occurred because of a physical change, is difficult, not to mention the inaccuracy of the resolution being detected. Sometimes resistance changes because of the heat emitted by ingredients due to the high number or extent of transformations. For this reason, developing a sensor that reacts extremely sensitively to temperature only and not pressure is key to developing a temperature measuring tactile sensor that mimics the skin.
There should not be any change in resistance that comes from external pressure, rubbing and or solvent (current flow).
Most conductive sensors claim that they measure heat and pressure at the same time because resistance changes according to heat, but this is not ideal. When change takes place in resistance, distinguishing the cause of the change, i.e., whether it occurred because of a physical change, is difficult, not to mention the inaccuracy of the resolution being detected. Sometimes resistance changes because of the heat emitted by ingredients due to the high number or extent of transformations. For this reason, developing a sensor that reacts extremely sensitively to temperature only and not pressure is key to developing a temperature measuring tactile sensor that mimics the skin.
(1) Optimizing the structure of the sensor that has a high sensitivity to temperature changes of less than 0.1℃
Because semi-crystalline polymers are characterized by their tendency to significantly change in volume as they approach melting point, they have high temperature detecting sensitivity when manufactured into a composite material. Therefore, we added polyethylene oxide (PEO), a semi-crystalline polymer, to the rGO/PVDF composite material in order to make a high-sensitivity temperature sensor (Fig 4).
Because semi-crystalline polymers are characterized by their tendency to significantly change in volume as they approach melting point, they have high temperature detecting sensitivity when manufactured into a composite material. Therefore, we added polyethylene oxide (PEO), a semi-crystalline polymer, to the rGO/PVDF composite material in order to make a high-sensitivity temperature sensor (Fig 4).
＜Figure 4. Photo and diagram of semi-crystalline polymer-based temperature sensor＞
The image on the right in Figure 4 illustrates the detection mechanism of a high-sensitivity temperature sensor. The semi-crystalline polymer-based temperature sensor will reduce in resistance due to the negative temperature coefficient (NTC) effect, which is unique to graphene (a thin membrane made of carbon atoms). In addition, temperature-induced expansion of semi-crystalline polymers narrows the gap between surrounding graphene, leading to a substantial change in the sensor’s resistance. A semi-crystalline polymer-based temperature sensor has approximately 15 times the temperature sensitivity of a non-semi-crystalline polymer-based temperature sensor. We confirmed that this high-sensitivity temperature sensor is able to detect in real-time even a slight change in temperature of less than 0.1℃ (Fig 5).
＜Figure 5. Characteristics of semi-crystalline polymer-based temperature sensors＞
(2) A sensor structure that does not change according to external pressure, rubbing and/or solvent (current flow).
When it comes to sensors that detect coolness that comes from applying a cosmetics product, it is critical to minimize reactions to tactile stimulations, such as pressure or bending. We confirmed that semi-crystalline polymer-based temperature sensors do not react to stimulations such as pressure or bending because we controlled the thickness to a very fine degree (~30μm) and implemented a co-planar electrode design (Fig 6).
When it comes to sensors that detect coolness that comes from applying a cosmetics product, it is critical to minimize reactions to tactile stimulations, such as pressure or bending. We confirmed that semi-crystalline polymer-based temperature sensors do not react to stimulations such as pressure or bending because we controlled the thickness to a very fine degree (~30μm) and implemented a co-planar electrode design (Fig 6).
＜Figure 6. Characteristics of sensors that do not change in temperature due to pressure or bending＞
Different components of cosmetics are responsible for the coolness we feel when we apply a product. Usually, however, the coolness comes from the solvent in the product. For this reason, being able to distinguish different solvents is a key requirement for a coolness sensor for cosmetics. Cosmetic solvents are often made up of water and ethanol, and when these solvents are dropped directly on the temperature sensor, it becomes difficult to measure the change in temperature due to the heat of vaporization coming from the current flow of solvents.
Therefore, the sensor needs to be equipped with a thin protective film to prevent direct contact from occurring between PDMS (a widely used silicon material) and the solvent as well as to detect the subtle heat of vaporization of the solvent. In the case of our sensor, we used DMS to coat the thin film (Fig 7).
Therefore, the sensor needs to be equipped with a thin protective film to prevent direct contact from occurring between PDMS (a widely used silicon material) and the solvent as well as to detect the subtle heat of vaporization of the solvent. In the case of our sensor, we used DMS to coat the thin film (Fig 7).
＜Figure 7. Protection film that is used to assess the solvent detection capability of the coolness sensor＞
We observed the change in temperature of water droplets of varying temperatures in order to assess the capability of the coolness sensor. We confirmed that the temperature climbed instantaneously and cooled naturally when 30, 40, 50℃ water droplets came in contact with the sensor (Fig 8).
＜Figure 8. Assessing the solvent detection capability of the coolness sensor＞
Different solvents have different heats of vaporization, which then leads to a difference in the degree of coolness. The coolness sensor we developed is able to detect and distinguish the degree of coolness that varies by solvent. (Fig 8). The graph on the right demonstrates the results of distinguished detection, showing that the and size of the peak vary according to heat of vaporization. We further confirmed that the time required for the resistance value to return to the default value also varied according to the speed of vaporization.
At present
Based on the structure of the sensor, which mimics the delicate tactile capabilities of human skin, and combined with Amorepacific R&I’s very own accumulated expertise and customer data, we were able to create a predictable algorithm. The findings of this co-research were published in the ACS Nano journal this February. Currently at Amorepacific R&I, the research is continuing with measurement facilities, so that we may be able to deliver the best customer experience possible using sensors that bring us closer to the sophisticated world of the sense of touch.
-
Like
1 -
Recommend
1 -
Thumbs up
1 -
Supporting
0 -
Want follow-up article
1
