JP7413953B2 - Charge sensor, robot hand, and sample determination device - Google Patents

Description

The present invention relates to an electrostatic charge sensor, a robot hand, and a sample determination device.

Conventionally, in order to stably grasp and identify fragile substances or substances with irregular shapes using a robot hand, attempts have been made to grasp substances of various shapes by making the exterior part of the robot hand deformable. ing. Such a robot hand is called a soft robot (for example, Patent Document 1).

In addition, for stable gripping and identification, a method is known in which thousands of electrical resistance pressure sensors are closely fixed on the surface of the robot hand and the contact pressure with the surface of the gripped object is analyzed using AI. (For example, Non-Patent Document 1). In addition, as a sensor used in a robot hand, it has been proposed to incorporate not only an electrical resistance pressure sensor but also a drive mechanism using deformation tension of dielectric elastomer, thread tension, pneumatic pressure, or hydraulic pressure (for example, Non-Patent Document 2 ).

US Patent No. 9464642

Nature562, 698-702 Journal of the Robotics Society, Volume 37, Issue 1, 2019

To control the robot hand, it is conceivable to use a camera or the like to recognize the shape of the object to be grasped from an image. In this case, it is required that there be no obstacles between the camera and the object to be grasped. Also, a considerable amount of learning is required to predict the hardness and amount of deformation from the image and have the device grasp it. With considerable study, it may be possible to predict approximate characteristics. However, it is difficult to predict the actual hardness and weight of a substance, and there is a risk that it may not be possible to grip the substance stably.

On the other hand, it is conceivable to provide a sensor on the surface of the robot hand and control the gripping position and gripping force using signals from the sensor. This method is effective when gripping something that has a certain degree of hardness. However, this method is not necessarily effective in cases where the cup itself is easily deformed, such as a side dish cup for a lunch box, and if the deformation is large, the contents may spill out. That is, due to the deformation of the cup, the pressure change due to contact with the cup is small. Also. If the position or force is reacted to the increase in pressure after a pressure change has occurred, the contents will not be constant (this is easier to understand if you imagine that the contents of a cup are an assortment of beans, etc.), so it is necessary to control it well. That's difficult. In addition, even if we try to respond to this slight pressure change, at the speed of the robot normally used, the end mechanism of the robot hand will move significantly between the time of contact and the time the robot determines that the robot has grasped the contents. Sometimes it spilled out. Reducing the speed of the robot hand to address this problem would undermine the purpose of using the robot, which is to perform large amounts of processing at high speed. For these reasons, it is currently not possible to introduce robots to the task of assembling food into flexible lunch containers and the like.

In other words, the conventional technology does not provide a robot that has a simple structure, is easily deformed, and is capable of properly gripping a container whose contents would spill if it were deformed. An object of the present invention is to provide a technique that makes it possible to more appropriately grasp a specimen.

As a result of intensive study, the inventors of the present application have devised a way to use a charge sensor that detects the presence or absence of contact based on a change in the capacitance of the charge sensor itself. I came to think of installing a sensor.

The present invention is as follows.
[1] A first charging section including a first electrode and a first dielectric;
a second charging section including a second electrode paired with the first electrode and a second dielectric;
an extraction unit that extracts a signal generated by a change in capacitance between the first charging unit and the second charging unit from the first electrode,
The first dielectric body is formed using an elastic member and includes a hollow portion that is hollow inside;
The hollow part includes an opening through which air enters and exits between the inside and outside as the hollow part deforms due to contact with the specimen,
The extraction section is a charging sensor that extracts a signal generated due to deformation of the hollow section from the first electrode.

[2] The first charging section includes the first dielectric, the first electrode, a noise removal layer that removes noise in the signal, and a third electrode that is grounded together with the second electrode. The charge sensor according to [1], which has a structure in which electrodes are stacked.

[3] The charge sensor according to [2], wherein the noise removal layer has a volume resistivity of 10 ⁵ Ω or more.

[4] The charging sensor according to [2] or [3], further including a switch that grounds the first electrode.

[5] The hollow portion is formed using at least one selected from silicone rubber, urethane rubber, acrylic rubber, natural rubber, and Teflon rubber, according to any one of [1] to [4]. Charge sensor.

[6] A robot hand including a first finger and a second finger different from the first finger,
The first charging portion described in any one of [1] to [5] is provided on the first finger, and the first charging portion described in any one of [1] to [5] is provided on the second finger. A robot hand characterized in that a second charging section is provided.

[7] The robot hand according to [6], in which the second finger is used as the second dielectric.

[8] The signal taken out from the take-out part depending on the state of the specimen held by the robot hand according to [6] or [7] and the hollow part that can be deformed by contacting the specimen. A specimen determination device, comprising a determination circuit that determines at least one of physical properties of the specimen and a state of the specimen with respect to the hollow portion, based on a shape.

[9] The specimen determining device according to [8], wherein the determining circuit determines the hardness of the specimen based on the magnitude of the level of the signal.

[10] According to [8] or [9], the determination circuit determines the elasticity of the specimen based on the decay time of the level of the signal accompanying the release of pressure supply from the specimen to the hollow portion. Determination device for the described specimen.

[11] The determination circuit according to any one of [8] to [10], wherein the determination circuit determines the size of the specimen based on the timing of a rise of the signal accompanying contact of the specimen with the hollow portion. Determination device for the described specimen.

According to the present invention, it is possible to more appropriately grasp a specimen.

FIG. 1 is a diagram showing an example of a charging sensor. An example of the configuration of the sensor main body portion shown in FIG. 1 is shown. FIG. 3 is a diagram showing an example of the configuration of a robot hand provided with the charging sensor shown in FIG. 1. FIG. 4 is a diagram showing the configuration of an apparatus including a charging sensor used in the demonstration experiment. FIG. 5 shows the results of the demonstration experiment. FIG. 6 is a diagram showing an example of a pressure waveform and a response signal obtained by pressure supply. FIG. 7 is a block diagram showing a configuration example of a sample determination device. FIG. 8 is a table showing an example of determination information. FIG. 9 is a flowchart illustrating an example of hardness and specimen type determination processing. FIG. 10 is a flowchart illustrating an example of sample determination processing.

The charging sensor according to the embodiment includes a first charging section that includes a first electrode and a first dielectric, and a second charging section that includes a second electrode and a second dielectric that are paired with the first electrode. The device includes two charging sections, and an extraction section that extracts a signal generated by a change in capacitance between the first charging section and the second charging section from the first electrode. The first dielectric body is formed using an elastic member and includes a hollow portion that is hollow inside. The hollow portion includes an opening through which air enters and exits between the inside and outside as the hollow portion deforms due to contact with the specimen. The extraction section extracts a signal generated due to the deformation of the hollow from the first electrode.

According to such a charge sensor, when a specimen approaches the charge sensor, a change in capacitance between the charge sensor and the specimen can be detected before contact occurs. As a result, the moving speed of the robot hand can be reduced before the specimen comes into contact with the charging sensor, based on the signal detected before the robot hand comes into contact with the specimen. As a result, even if the specimen is easily deformed, the gripping operation can be performed without significantly deforming the specimen or spilling the contents.

Further, the hollow part is provided with an air inlet/outlet that communicates with the outside and allows the air inside the hollow part to escape from the hollow part after it has come into contact with the sample. This makes it possible to continuously and significantly detect changes in capacitance even after contact. Since a signal can be obtained before contact with the specimen, and a signal corresponding to a change in the contact state can be obtained even after contact, it is possible to appropriately grasp the specimen.

This makes it possible to detect the presence of a specimen before contact, to recognize contact even with a weak contact pressure, and to obtain a signal that corresponds to a wide range of contact pressure. It is possible to provide a determination device that determines a specimen using a robot hand provided with a charge sensor and a charge sensor.

(sample)
The "specimen" is an object (solid) with which the contact surface of the sensor comes into contact, and there are no restrictions on its material or shape. Examples of specimens include plates, blocks, flower pots, sheets, tableware, desks, containers, gloves, bags, etc. made of metal plates such as aluminum, zinc, copper, or steel, paper, rubber, plastic film, or glass. Examples include products, foods such as mandarin oranges, bananas, rice, bread, konjac, and meat, insects, animals, and fish, and at least one of these examples may be selected as the specimen.

(Charge sensor)
Charge sensors do not require a power source and have a zero-based signal strength (level) when not sensing. Therefore, the SN ratio (sensitivity) when the signal output from the charge sensor is amplified by the amplifier is high. Therefore, the response to vibration can be detected with high accuracy from the output signal of the charging sensor.

In this specification, a "charged sensor" is a sensor that utilizes a charging principle such as frictional power generation or electret. Charge sensors generate their own voltage by utilizing the principle of charge generation. Therefore, a light emitting device, a power source, etc. are not required. Therefore, the structure is simpler and less susceptible to external disturbances than an optical fiber sensor that detects a response based on constant light or fluctuations in electrical signals.

The charging sensor can have a triboelectric charging structure or an electret charging structure. The triboelectric charging structure has a structure in which two (pair of) charging parts face each other, that is, two charging parts, each including an electrode and a dielectric material, are stacked (opposed) with a gap therebetween. The electret charging structure has a structure in which a dielectric material having a plurality of positively and negatively charged bubbles is sandwiched between electrodes, which is a set of charging portions consisting of an electrode and a dielectric material. A charging sensor having a triboelectric charging structure is referred to as a tribocharging sensor, and a charging sensor having an electret charging structure is referred to as an electret charging sensor.

The advantage of using a charged sensor is that it can detect much weaker force than a sensor that uses the piezo effect, and the waveform obtained differs depending on the physical properties of the material it comes in contact with and how easily it deforms. be.

(dielectric)
The dielectric material that can be applied to the first dielectric and the second dielectric is not particularly limited as long as it has chargeability. Examples include polymers (resins) and nonmetallic substances. Polymers include silicone resins (polydimethylsiloxane (PDMS), etc.), fluororesins (polytetrafluoroethylene (PTFE), etc.), polyimide, polyvinyl chloride, polystyrene, polyolefins (polyethylene, polypropylene, etc.), polyester resins (polyethylene terephthalate, etc.). (PET), etc.), polycarbonate resins, acrylic resins (polymethyl methacrylate (PMMA), etc.), polyamide resins (nylon, etc.), and cellulose. Examples of the nonmetallic substance include oxides such as silica and alumina. The dielectric material is preferably formed of a resin film or a thin film of a non-metallic material in order to make the hollow charge sensor smaller and thinner.

The dielectric constant (εDI) of the dielectric is preferably 5 or more, more preferably 8 or more. Furthermore, although there is no upper limit to the dielectric constant of the dielectric, a higher one is preferable. The relative dielectric constant can be calculated from the following formula 1 by forming electrodes on both sides of the layer of the material to be measured (sample) to create a parallel plate capacitor, and measuring the capacitance C of the capacitor. .
C=ε0・εr・A/d... (Formula 1)
Here, εr is the relative permittivity, ε0 is the permittivity of vacuum, A is the area of the capacitor, and d is the thickness of the layer of the material to be measured. The capacitance C of the capacitor can be measured using an LCR meter, an impedance analyzer, or the like.

It is preferable to use the following materials as the outermost layer of the portion of the dielectric that causes friction due to contact. That is, examples of the material for the outermost layer include polymers, nonmetallic substances, and metallic substances. Examples of the polymer include those mentioned above and melamine resin. Further, examples of non-metallic substances include oxides such as silica and alumina. Further, examples of the metal substance include aluminum, iron, nickel, silver, gold, platinum, copper, chromium, titanium, molybdenum, indium metal, and alloys of these metals.

(electrode)
The electrodes that can be used as the first to third electrodes are not particularly limited as long as they have high electrical conductivity, and any material can be suitably used. The electrode is, for example, aluminum, iron, nickel, silver, gold, platinum, copper, chromium, titanium, molybdenum, or indium metal, and alloys of these metals can also be used. The electrodes do not need to be made of metal as long as they are made of a material that has high electrical conductivity. For the electrode, for example, a semiconductor material such as doped silicon (Si), a metal oxide such as indium tin oxide (ITO), or a conductive polymer such as PEDOT-PSS can also be used. The electrode is preferably made of a material that is soft, deformable, and resistant to damage, and is preferably as thin as practical, usually 100 μm or less, more preferably 50 μm or less, most preferably 25 μm or less, and the lower limit The average diameter is 10 μm or more. By making it thin like this, it is possible to prevent malfunctions due to damage to the electrodes without reducing the sensitivity of the charge sensor.

(Removal section)
The extraction section receives a response signal (a signal generated by a change in capacitance between the first charging section and the second charging section) from a first electrode of the charging sensor, and a deformation of the hollow section due to contact with the specimen. This is to extract the signals generated in conjunction with the When the response signal is an electrical signal, it is constructed using a conductor. For example, the extraction part is a lead wire, a terminal, a connector, etc. electrically connected to the electrode.

(robot hand)
One embodiment of the invention includes a robotic hand. The robot hand includes a plurality of fingers including, for example, a first finger and a second finger different from the first finger. A second charging section is provided. A second finger may be used as the second dielectric. In the case where "the second charged part is provided on the second finger", in addition to the case where the second electrode included in the second charged part is physically attached to the second finger, This includes a case where the finger is electrically connected to the second finger at a position separated from the second finger.

A robot hand to which the charged sensor according to the present invention is applied has a plurality of fingers, at least one of which is provided with a sensor main body portion including a first charged portion, and is electrically connected to or connected to the other fingers. At least one of the sections can have a counter electrode. The number of fingers may be 2 or more, for example, 2-6. However, there is no upper limit.

The fingers are made of an elastic member. Preferably, the exterior of the finger has softness. However, it is not essential to use elastic members or soft materials. The fingers are driven by the actuator and grip the specimen.

In consideration of the case where the specimen to be gripped or held by the robot hand is soft and easily deformed, the outer surfaces of the fingers are preferably made of a soft material. For this reason, it is preferable that the robot hand has an internal hollow part, is made of an elastic or soft thin film member, and has an air suction port for sucking the air inside the hollow part. The inner hollow part of the grip part swells in the atmosphere, and the distance between the tips of the fingers becomes wide. However, when the inner hollow part contracts at a low pressure due to suction, for example, with a force of about -35 Kpa, each of the fingers expands. The tip approaches and closes to grasp the specimen. The grip part and the arm part are movable in the vertical direction and the horizontal direction (XYZ directions of the three-axis coordinate system), and the inclination with respect to the vertical axis can be changed.

(Sample determination device)
One embodiment of the present invention includes a sample determination device. The determination device determines the physical properties of the specimen based on the shape of the signal taken out from the extraction part depending on the state of the specimen gripped by a robot hand equipped with an electrostatic charge sensor and the hollow part that can be deformed by contacting the specimen. , and a determination circuit that determines at least one of the conditions of the hollow portion of the specimen. It is preferable that the determination circuit employs a configuration that determines the hardness of the specimen based on the signal level and magnitude of the signal. Further, it is preferable that the determination circuit employs a configuration in which the elasticity of the specimen is determined based on the decay time of the signal level accompanying the release of pressure supply from the specimen to the hollow portion. Further, it is preferable that the determination circuit employs a configuration in which the size of the specimen is determined based on the timing of the rise of a signal accompanying contact of the specimen with the hollow portion.

The present invention will be explained in detail below. The description given below is one embodiment of the present invention, and the present invention is not limited to these contents unless it exceeds the gist thereof.

<Charge sensor>
FIG. 1 is a diagram showing an example of the configuration of a charging sensor according to an embodiment of the present invention. FIG. 2 shows an example of the configuration of the sensor body. FIG. 3 schematically shows a case where the charging sensor shown in FIG. 1 is attached to a robot hand.

In FIG. 1, the charging sensor 41 is roughly divided into two parts. One is a sensor body portion 10 having a hollow portion 44, which is provided in a portion of the soft robot that grips the specimen 60 (robot hand (also referred to as gripping portion); see FIG. 2). The other is a counter electrode section 20 that exists independently from the sensor main body section 10.

The sensor body portion 10 includes a support 54A, an electrode 532B (corresponding to a third electrode) formed on the support 54A, and a high resistance layer (corresponding to a noise removal layer) formed on the electrode 532B. 55A, and a charging section 51A (corresponding to a first charging section) provided on the high resistance layer 55A.

The charging section 51A includes an electrode 53A (corresponding to a first electrode) formed on a high-resistance layer 55A, and a dielectric 52A (corresponding to a first dielectric) formed on the electrode 53A. . The dielectric 52A has a dome-shaped wall portion 58 and includes a hollow portion 44 in which the interior 45 is hollow. The hollow part 44 includes an air inlet/outlet 46 that communicates the inside 45 of the hollow part 44 with the outside. When pressure is applied from the outside, the hollow part 44 deforms (collapses) and releases the air in the interior 45 to the outside through the inlet/outlet 46, and then, when the pressure decreases, air is taken in from the outside through the inlet/outlet 46 and returns to its original shape. Return to There may be two or more entrances and exits 46, and the entrance and exit may be separate.

The wiring connected to the electrode 532B of the sensor body portion 10 and the wiring connected to the electrode 531B of the counter electrode portion 20 are connected (conducted) to the lead wire 43A and are configured to have the same potential. Electrode 532B and electrode 531B are grounded through lead wire 43A.

A switch 532C is provided between the wiring drawn out from the electrode 532 and the lead wire 43B connected to the electrode 53A. When the switch 532C is turned on (closed), the potential of the electrode 53A becomes grounded. The lead wire 43B converts the current or voltage generated by a change in capacitance between the charged portion 51A and the charged portion 51B, and the current or voltage generated by the deformation of the hollow portion 44 due to contact with the specimen 60 into a sensor signal (signal). ) acts as a take-out unit to take out (detect).

The counter electrode section 20 includes a charging section 51B. The charging section 51B has, for example, a structure in which a support 54B, an electrode 531B, and a dielectric 52B are laminated. Note that the dielectric 52B is provided as necessary.

Each of the charging parts 51A and 51B is formed into a rectangular shape with a plane of 18 mm x 18 mm. However, the hollow portion 44 is formed in a planar circular shape. The support body 54A is formed using a polyimide film with a thickness of 30 μm. The electrode 53A is a 100 nm thick copper film, and the electrode 532B is a 100 nm thick aluminum film. Electrode 532B is formed by vapor deposition onto support 54A, which is a 30 μm thick polyimide film.

The electrode 531B is a 100 nm thick aluminum film deposited on a support 54B made of a 16 μm thick PET film. In the example shown in FIG. 3, the dielectric 52B is omitted and the finger 2B acts as the dielectric 52B.

The electrodes 531B and 532B are connected to a conducting wire such as an aluminum thin film or a copper thin film as a wiring, and are connected to a lead wire 43A as a ground wire. Connecting conductors is preferable because it is less susceptible to environmental changes such as humidity and temperature. The support body 54B is formed using a PET film having a thickness of 25 μm.

When the electrode 531B is a dielectric material, the electrode 531B may also serve as the dielectric 52B. Preferred materials for the dielectrics 52B and 52A are resin materials having elasticity, such as silicone rubber, urethane rubber, acrylic rubber, natural rubber, and Teflon rubber. Particularly preferred are silicone resins, urethane resins, acrylic resins, and epoxy resins.

The lead wires 43A and 43B are connected to, for example, an electrometer, and the response signal waveform displayed on the display of the electrometer is observed. The support 54B acts as a protective layer to prevent damage to the dielectric 52B. In order to increase the sensitivity of the charge sensor 41, the support body 54B is preferably made of a member that is easily deformed and that is less likely to cause friction. It is preferable that the electrodes 53A and 532B are made of a material that is easily deformable and less likely to be damaged by deformation.

One of the dielectrics 52A and 52B can be made of a material that easily stores charges, and the other can be made of a material that does not easily store charges. Further, the dielectric 52A and the dielectric 52B may be made of materials having similar characteristics. When attaching the sensor main body portion 10 to the finger 2A by adhering the support 54A to the finger 2A, it is preferable to use a material for the support 54A that is less likely to be damaged and has a high affinity with adhesives.

The charging section 51A and the charging section 51B can each become a charging section with positive and negative charges when performing frictional power generation. In this case, it does not matter whether the electrode 53A or the electrode 53B is positive or negative. For example, the electrode 53A may be positively charged and the electrode 53B may be negatively charged. Conversely, the electrode 53A may be negatively charged and the electrode 53B may be positively charged.

The hollow portion 44 included in the dielectric 52A has a dome shape in FIG. 1 . However, it is preferable that the hollow part 44 has a dome shape such as a hemisphere, a cone, or a 4 to 20 polyhedron with length and width of 5 mm to 8 mm and height of 1 mm to 10 mm. The thickness between the inside and outside of the hollow portion 44 (thickness of the wall portion 58) is preferably from 0.01 mm to 5 mm. As the material for the hollow portion 44, it is preferable to use an elastic resin material such as silicone rubber, urethane rubber, acrylic rubber, natural rubber, or Teflon rubber. Particularly preferably, silicone resin, urethane resin, or acrylic resin is used as the material for the hollow portion 44.

Further, the hollow portion 44 and the dielectric body 52A may be made of the same material or may be partially made of a plurality of different materials. When different materials are used, it is preferable to select materials that are widely separated on the charging array between the side in contact with the specimen 60 and the side in contact with the electrode 53A. For example, a material that is easily charged negatively, such as silicone rubber, is selected for the side that contacts the specimen 60, and a material that is easily charged positively, such as nylon resin, is selected for the side that is close to the electrode 53A.

There is no particular restriction on the shape of the entrance/exit 46 of the hollow portion 44, and a circular hole, slit, or the like may be adopted. The entrance/exit 46 is closed when no pressure is applied to the hollow section 44 from the specimen 60 or the like, but may open when the wall section 58 is deformed by external pressure.

Note that the thickness and material of each of the dielectric, electrode, and support included in the charging sensor 41 can be set as appropriate. The preferred film thicknesses of the dielectric, electrode, and support are in the ranges of 0.01 μm to 500 μm, 0.01 μm to 1000 μm, and 1 μm to 1000 μm, respectively. More preferably, the range is from 0.05 μm to 200 μm, from 0.02 μm to 500 μm, and from 5 μm to 500 μm.

The high resistance layer 55A sandwiched between the electrode 53A and the electrode 532B has a higher volume resistivity than the electrode 53A and the electrode 532A. The material of the high resistance layer 55A can be appropriately selected from organic or inorganic materials having a volume resistivity of preferably 10 ⁵ Ωcm or more and less than 10 ¹² Ωcm, more preferably 10 ⁶ Ωcm or more and less than 10 ¹² Ωcm. The high resistance layer 55A has a function of suppressing high frequency component noise and baseline fluctuations and stabilizing the sensor signal. As the material for the high resistance layer 55A, it is preferable to use an elastic resin material such as paper, polyimide resin, silicone rubber, urethane rubber, acrylic rubber, natural rubber, or Teflon rubber. In particular, it is preferable to use silicone resin, urethane resin, and acrylic resin.

The preferred thickness of the high resistance layer 55A is 10 nm to 5 mm, more preferably 100 nm to 500 μm. The sensor signal stabilizing effect of the high resistance layer 55A is significantly higher than the frequency of the detected signal. This is thought to be because charges of high frequency components easily pass through the high resistance layer 55A. When the volume resistivity is significantly lower than 10 ⁵ Ωcm, the signal intensity decreases, and when it is significantly higher than 10 ¹² Ωcm, the signal stabilizing effect is reduced. The volume resistivity is preferably measured by attaching 35 μm copper foil to both sides of a 5 mm x 5 mm square organic resin layer sample to create a bonded volume resistivity sample. As a measuring device, for example, a high resistance tester TM385 manufactured by Tecman can be used for measurement.

<Robot hand>
When the charging sensor 41 is installed on the robot hand (grip section) 2, as shown in FIG. The portion 10 is installed, and the counter electrode section 20 is installed on a finger different from the finger 2A (second finger (finger 2B)).

As shown in FIG. 2, the sensor body portion 10 is placed on the finger 2A. The robot hand 2 opens and closes the fingers 2A and 2B, and during the closing operation, closes the fingers 2A and 2B so that the pads at the tips of the fingers come into contact with each other, and holds the specimen 60. Sandwich. At this time, the pressing surface (lower surface in FIG. 2) 59 of the specimen 60 comes into contact with the contact surface that is the surface of the sensor main body portion 10 provided on the finger 2A, that is, the wall portion 58 of the hollow portion 44. The sensor main body portion 10 detects a change in the amount of charge (charge amount) using a lead wire (takeout portion) 43B connected to the electrode 53A. The robot hand 2 has an internal hollow part 21 and an air suction port 22 for sucking air inside the internal hollow part 21. In the robot hand 2, the internal hollow portion 21 expands in the atmosphere, and the distance between the tips of the fingers 2A and 2B becomes wide. On the other hand, when the internal hollow part 22 contracts under reduced pressure by suctioning the air in the internal hollow part 21 from the air suction port 22 with a force of about -35 Kpa, for example, the tips of the fingers 2A and 2B approach each other. , close and grasp the specimen 60.

On the other hand, in FIG. 3, the counter electrode section 20 does not have the dielectric 52B unlike in FIG. 1, and the finger 2B acts as the dielectric 52B. As long as the finger 2B is electrically connected to the electrode 531B, the counter electrode section 20 does not need to be on the finger 2B. By providing this grounded counter electrode section 20, not only the tactile detection that senses the pressure caused by the contact between the sensor body section 10 and the specimen 60 but also the electrostatic charge field between the sensor body section 10 and the specimen 60 can be avoided. It can also have the function of detecting in close contact state.

In the charged sensor 41 shown in FIGS. 1 and 3, the charged portion 52A of the sensor main body portion 10 connects a hollow portion 44 in which a hollow interior 45 communicates with the outside (outside space) via an air inlet/outlet 46. have. The wall portion 58 of the hollow portion 44 is made of a flexible (elastic) material. The wall portion 58 is easily deformable against the weak contact pressure of the specimen 60. Further, when external pressure is generated due to contact with the specimen 60, the hollow portion 44 can release air from the interior 45 through the entrance/exit port 46. If the hollow part 44 does not have the entrance/exit 46, the air pressure in the interior 45 will increase due to the deformation, and the amount of deformation of the charging part 52A will be small with respect to the increase in pressure, resulting in a decrease in sensitivity. By allowing air to escape through the entrance/exit port 46, a decrease in sensitivity can be suppressed.

When the specimen 60 is held between the fingers 2A and 2B, the capacitance (capacitance C) between the electrodes 531B and 53A of the charging sensor 41 changes as the fingers 2A and 2B of the robot hand 2 approach the specimen 60. . Further, when the specimen 60 and the charging portion 52A (hollow portion 44) come into contact, charges move from the specimen 60 to the charging portion 52A, and charging power generation occurs. When the gripping force of the fingers 2A and 2B is further strengthened, the air in the interior 45 of the hollow portion 44 escapes from the opening 46 due to the pressure from the specimen 60, and the hollow portion 44 is collapsed. If the pressurization continues even after most of the air has escaped from the interior 45, the wall 58 will continue to contract and deform due to its flexibility. In this manner, while the hollow portion 44 is deformed, crushed, and further compressed, charging and power generation continues to occur. Due to the charging power generation, the voltage between the electrode 532B and the electrode 53A changes, and this change is detected as a sensor signal from the extraction part (lead wire connected to the electrode 53A) 43B.

According to the charging sensor 41, due to the effect of the hollow portion 44 described above, pressure from a wide range of specimens can be detected, from a weak tactile pressure of 1 g or less to a strong pressure change of several hundred g. As described later, the charge sensor 41 can detect differences in the size and physical properties of various specimens as changes in the sensor signal by using the first to third detection modes, making it possible to highly identify specimens. .

<Proof experiment>
FIG. 4 is an explanatory diagram of the apparatus used to obtain the results of the demonstration experiment shown in FIG. 5. The sensor main body part 10 is installed on a flat surface, and the robot hand (gripping part) 200 that grips (pinches) the side surface of the specimen 60 (a Lego block manufactured by Lego Corporation) is brought close to the sensor main body part 10. The robot hand 200 has fingers 2C and 2D, and uses an opening/closing mechanism 25 using a pinion and a rack to widen or narrow the distance between the fingers 2C and 2D. It can be opened and closed. The robot hand 200 can be moved up and down by a vertical movement mechanism 23 using a pinion and a rack mechanism.

A counter electrode section 20 is provided on the upper part of the robot hand 200. The robot hand 200 is a scalar robot "VS-ASR" manufactured by Vstone. By operating the vertical movement mechanism 23, the specimen 60 gripped by the robot hand 200 is lowered stepwise toward the sensor body portion 10, brought into contact with the sensor body portion 10, and removed from the electrode 53A (lead wire 43B). I looked at the change in the signal from the sensor (sensor signal).

FIG. 5 shows the measurement results of the sensor signals. The upper graph in FIG. 5 shows the record of the vertical movement of the lower end of the specimen 60 gripped by the robot hand 200. The lower graph in FIG. 5 shows a signal indicating the pressure on the charging sensor 41 (output signal of the load cell), and an output signal (sensor signal) of the charging sensor 41 due to changes in capacitance and pressure changes as the specimen 60 advances and retreats. and

The plane on which the sensor body part 10 is placed is on the stand of a load cell type precision balance made by PRC, and the change in load when the specimen 60 and the sensor body part 10 come into contact is measured using the output signal of the load cell of the precision balance. saw.

In the experiment, the initial state was set so that the gap between the specimen 60 and the hollow part 44 was 3.0 mm, and this initial state was set as the starting point (0 mm: the 0 mm point on the graph in Figure 5) ((a in Figure 5). )reference).

Next, the robot hand 200 was driven to lower the specimen 60 by 11.8 mm toward the hollow portion 44 (downward) over 3.8 seconds at a constant speed. During this movement, the specimen 60 descended 3.0 mm 1.0 seconds after starting to descend, and came into contact with the surface of the hollow portion 44 (see (b) in FIG. 5).

Looking at the sensor signal intensity shown in the lower graph, a sensor signal is already obtained before the sensor descends 3 mm and contacts the sensor body (see (1) in FIG. 5). This can be seen from the fact that the load obtained by the load cell provided at the lower part of the sensor main body portion 10 is zero, indicating that the approach of the specimen 60 can be detected at the time of non-contact.

When the hollow portion 44 (sensor body portion 10) and the specimen 60 come into contact, a load is detected by the load cell (see the graph of “Load cell”). After that, the specimen 60 descends further by 5.3 mm over 1.7 seconds, continues to push the hollow part 44, and the interior 45 of the hollow part 44 is contracted and compressed, and finally the hollow part 44 is almost collapsed (the hollow layer (see Figure 5(c)).

It can be seen that the signal (sensor signal) from the charging sensor 41 up to this point increases almost linearly as the pressure increases (see (1) and (2) in FIG. 5). This can be seen from the fact that the waveform of the output signal of the load cell and the waveform of the sensor signal are almost parallel. It can be seen that the air in the hollow part 44 comes out through the entrance/exit 46, and the hollow part 44 is in a substantially collapsed state, and the deformation due to the compression of the air in the inside 45 is not suppressed.

Next, when the collapsed hollow part 44 begins to be further compressed by the pressure of the specimen 60, the slope of the sensor signal changes, but the slope itself becomes constant (see (3) in FIG. 5). The slope of the output signal of the load cell changed greatly due to the depression of the hollow layer ((c) in FIG. 5), indicating that a large shape change occurred due to the contraction of the hollow part 44.

The specimen 60 further descended by 3.5 mm over an additional 1.1 seconds, continued to compress the contracted hollow charged part with even stronger pressure, and stopped (concave layer compression: see (d) in FIG. 5). The signal from the load signal at this time indicates the maximum pressure value, and it can be seen that the specimen 60 is applying the maximum pressure to the hollow portion 44 (sensor main body portion 10).

Next, the position of the specimen 60 was held for 2.1 seconds ((3) in FIG. 5). The hollow portion 44, which has been temporarily compressed to the maximum within this time, undergoes deformation (tries to expand) to relax its contracted state in order to return to a steady state. Therefore, a change in capacitance occurred due to a change in the shape of the hollow portion 44, and this change was detected as an increase in the sensor signal. On the other hand, the signal from the load cell detected a decrease in pressure.

Thereafter, the specimen 60 was raised 11.8 mm over 1.1 seconds and returned to the initial position (specimen detachment: (e) in FIG. 5). As the specimen 60 rises, the specimen 60 separates from the surface of the hollow part 44, and the decrease in capacitance associated with this separation reduces the intensity of the sensor signal, and the intensity of the signal from the load cell pushes against the hollow part 44. decreased due to decrease in pressure.

Even after the specimen 60 returns to its original position and the hollow portion 44 is restored to its original state, the charge remains in the hollow portion 44. It showed a lower voltage (-0.94 V) than the sensor signal value (initial signal voltage) at the start of the experiment. This residual charge is also considered to be related to the charging characteristics of the specimen 60. This residual charge is. It was removed by turning on (closed) switch 532C and shorting between lead wires 43A and 43B, and the sensor signal showed the initial voltage before the experiment.

From the evaluation results of the sensor characteristics of the charging sensor 41, the charging sensor 41 is in a non-contact specimen electric field detection mode (first detection mode: (1) in FIG. 5) before the specimen 60 contacts the surface of the hollow portion 44. , a low pressurization detection mode using the specimen 60 after contact (second detection mode: (2) in FIG. 5), and a high pressurization detection mode using the specimen 60 (third detection mode: (3) in FIG. 5). ). For any of the first to third detection modes, a good correlation was obtained with the hand setting position of the control program of the robot hand 200 and the pressure [g] measured by the load cell.

From these results, when the robot hand 200 grips the specimen 60, it becomes possible to detect the specimen 60 in a non-contact state and a wide pressure range from weak pressure to high pressure from the specimen 60. Therefore, it is possible to highly classify the specimen 60 to be grasped depending on its physical properties such as shape, softness, and elasticity.

<Example>
FIG. 6 shows the detection characteristics evaluation results of the charge sensor 41 attached to the finger of a robot hand (air vacuum type soft hand piSOFTGRIP made by Piab) attached to a robot arm (Dobot Magician) made by Shenzhen Yuejiang Technology. In the graph of FIG. 6, the vertical axis represents the sensor signal (response signal: the unit is voltage V) of the charging sensor 41, and the horizontal axis represents time (unit: seconds).

The evaluation results shown in FIG. 6 are based on the charge sensor 41 obtained when the robot hand 2 grips the specimen 60 for the purpose of evaluating the characteristics of the specimen 60 with different shapes and materials gripped by the robot hand 2. It was confirmed that the correlation between the sensor signal shape (waveform) from the sample 60, the size of the specimen 60, and the physical properties could be detected.

The specimens 60 used were garlic (weight 24 g, diameter approximately 4 cm, including garlic body and paper-like skin) and grape grains (weight 12 g, diameter 2.5 cm, soft and highly elastic).
We used an ABS ball (weight 3 g, diameter 2.5 cm, same size as a grape but hard) that was made using a 3D printer from ABS filament and was approximately the same size as a grape.

The robot hand 2 has three fingers, and one of the fingers is provided with the sensor body portion 10 (charged portion 51A), and another finger is provided with the counter electrode portion 20 (charged portion 51B). . The robot arm was operated so that the specimen 60 was located at the center of these three fingers, and the robot hand was fixed. Next, lead wires 43A and 43B of charge sensor 41 were short-circuited (switch 532C was turned on), and the sensor signal obtained from electrode 53A was set to 0V. Thereafter, the switch 532C is turned off, the air inside the internal hollow part 21 is sucked from the suction port 22 of the robot hand 2 at -35 Kps, and the fingers 2A and 2B are closed to grasp the specimen 60 for about 0.5 seconds. The suction state was maintained for 1.6 seconds.

The sensor signal (vertical axis, unit V) shown in the graph of FIG. 6 increases as the specimen 60 contacts and pressurizes the hollow part 44 at the start of grasping, and reaches the maximum pressurized state (peak of the sensor signal). reach. At this time, the hollow portion 44 temporarily enters the maximum contraction state. Thereafter, the pressure on the hollow portion 44 is gradually relieved and the hollow portion 44 expands until it reaches a steady state. From this, a signal with an attenuation curve was obtained after the signal peak.

Comparing the shapes of the sensor signals of the three types of specimens 60, garlic is larger than grapes and ABS balls, so it reaches the contact position with hollow portion 44 earlier than grapes and ABS balls. For this reason, the rising timing and peak of the sensor signal related to garlic (graph A) was observed earliest among the three types.

Also, when comparing the rise and decay curve of the sensor signal of graph A related to garlic with graph B of grapes and graph C of ABS ball, the decay curve of graph A is not a simple curve, but has multiple curves. A charged power generation component is observed. This is thought to be due to the frictional electrification that occurs when the garlic body and the electrically charged skin surrounding it come into contact with the hollow portion 44, causing a plurality of rising and decaying signals. This shows a good correlation between the complex garlic shape and the sensor signal shape.

On the other hand, graph B related to grapes and graph C related to ABS balls having approximately the same size and shape as grapes showed similar signal shapes (signal waveforms). However, for grapes, higher peaks and faster signal decay were observed compared to ABS balls. This is because grapes are more flexible and have higher elasticity than ABS balls, and tend to undergo large shape changes faster than ABS balls when pressurized. It shows a correlation.

The above results revealed that the hollow charge sensor of the present invention having multiple detection modes has high discrimination characteristics for specimens having different shapes, softness, elasticity, etc.

<Judgment device>
FIG. 8 is a block diagram showing a configuration example of a sample determination device applicable to the robot hand 2 including the charging sensor 41 according to the embodiment. The determination device 200 includes a processor 201, a storage device 202, a communication interface (communication IF) 203, an input device 204, and a display 205, which are interconnected via a bus 210.

The bus 210 is connected to an electromagnetic drive device (motor, solenoid, etc.) 206, and the electromagnetic drive device 206 is connected to a manipulator 207 including a robot hand (grip portion) 2. A hollow charge sensor 40 is attached to the robot hand 2, and the charge sensor 41 as a vibration responsive sensor is connected to the electrometer 2 via lead wires 43A and 43B.
Connected to 08.

As the determination device 200, a general-purpose or dedicated computer can be applied. Storage device 202 includes a main storage device and an auxiliary storage device. The main storage device is used as a storage area for programs and data, a work area for the processor 201, and a buffer area for communication data. The auxiliary storage device is used as a storage area for programs and data. The main memory is RAM (Random Access Memory) or a combination of RAM and ROM (Read Only Memory). Auxiliary storage devices include hard disks, SSDs (Solid State Drives), and EEPROs.
M, etc.

The communication IF 203 is an interface that performs information communication (input/output) with a server, other devices, etc. via a wired network or a wireless network. The communication IF 203 is, for example, a network interface card, a wireless communication module (circuit chip), or the like. The input device 204 is a key, button, pointing device, touch panel, or the like, and is used to input data or information to the determination device. Display 205 is used to display data and information.

The processor 201 is, for example, a CPU (Central Processing Unit), and the storage device 2
By executing the program stored in 02, it operates as a control device that controls the operation of the robot hand (grip section) 2. The processor 201 performs various processes based on commands received from an external device via the communication IF 203, commands input by an operator or the like using the input device 204, or commands generated by executing a program.

For example, the processor 201 calculates the control amount of the electromagnetic drive device 206 that controls the position, tilt, opening and closing of the fingers 2A and 2B, etc. of the robot hand 2, and supplies the electromagnetic drive device 206 with a control signal indicating the control amount. The electromagnetic drive device 206 includes a plurality of joints and linear sliders included in the manipulator 207, a plurality of stepping motors, servo motors, and solenoids corresponding to pumps and valves for adjusting the air pressure of the air actuator, and controls the amount of control given by the processor 201. Drive control is performed only by the amount according to (the direction and speed of rotation or movement). As a result, the position and inclination of the robot hand 2, opening and closing of the plurality of fingers included in the robot hand 2, etc. are performed.

A charging sensor 41 having a hollow portion 44 is attached to the robot hand 2 . Charge sensor 41 is connected to electrometer 208 via lead wires 43A and 43B. When the robot hand 2 grips the specimen 60 and the specimen 60 approaches the hollow portion 44, a potential difference occurs due to a change in the capacitance C of the hollow portion 44 (charging portion 52A). Further, when the specimen 60 contacts and pressurizes the hollow portion 44, a potential difference is generated due to frictional power generation. A response signal indicating a current or a temporal change in the current caused by these potential differences is input from the electrode 53A to the electrometer 208 through the lead wire 43B.

The electrometer 208 measures a response signal (temporal change in the response signal (current or voltage) due to repetition of pressure supply and release), and displays the information shown in FIGS. 5 and 6 on the display included in the electrometer 208. Displays a current or voltage waveform indicating a response signal (sensor signal), such as a sensor signal. By observing the waveform, the hardness, elasticity, shape, etc. of the specimen 60 can be visually determined.

(Processing example 1)
The response signal output from the charge sensor 41 is amplified by an amplifier included in the determination device 200 , converted into a digital signal by an AD converter, and then input to the processor 201 via the bus 210 . The digital signal input to the processor 210 indicates the output signal from the charging sensor 41. It is preferable to use a voltage waveform because the voltage waveform changes more than the current waveform.

For example, the processor 201 calculates (determines) the hardness of the specimen 60 gripped by the robot hand 2 by comparing data regarding various waveforms stored in advance in the storage device 202 and the waveform acquired from the charging sensor 41. )do. The storage device 202 stores data regarding waveforms depending on the hardness of the specimen 60 as determination information. Therefore, the processor 201 compares data regarding the waveform stored in the storage device 202 with the waveform acquired from the charging sensor 41 (referred to as "measurement value"), thereby determining the value of the specimen 60 gripped by the robot hand 2. Hardness can be calculated (determined).

FIG. 8 is a table showing an example of determination information. The table consists of one or more entries (records) including information indicating hardness, threshold value, and specimen type. The hardness may be any physical quantity or a value indicating relative hardness in comparison with a certain object. For example, it may be a relative value indicating the hardness between multiple types of specimens. In the example of the table shown in FIG. 8, the relationship between the hardness of specimens A, B, and C having different hardness and the determination threshold value is recorded for each record. A hardness of "1" is the softest, and the larger the value, the harder it is. The threshold value is TH1<TH2<TH3.

FIG. 9 is a flowchart illustrating an example of hardness and sample type determination processing performed by the processor 201. In step S01, the processor 201 grips a certain specimen 60 and acquires the measured value of the response signal output from the charging sensor 41. The measured value is, for example, a value indicating the amount of rise (increase amount) in the voltage level due to the supply of pressure, but other values may be used. As shown in graphs B and C of FIG. 6, when the two specimens 60 have the same size, the peak of the response signal of the softer specimen 60 becomes larger. As an example, the threshold may be a threshold for a peak of the response signal.

In step S02, the processor 201 determines whether the measured value belongs to a predetermined threshold range. The threshold range in step S02 is "0<measured value=<threshold TH1". If it is determined that the measured value falls within the threshold range (YES in step S02), the processor 201 determines that the hardness of the specimen 60 is “1” and that the specimen 60 is the specimen “A”. If it is determined in step S02 that the measured value does not fall within the threshold range (NO in step S02), the process proceeds to step S03.

In step S03, the processor 201 determines whether the measured value belongs to a predetermined threshold range. The threshold range in step S03 is "TH1<measured value=<threshold TH2". If it is determined that the measured value falls within the threshold range (YES in S03), the processor 201 determines that the hardness of the specimen 60 is “2” and that the specimen 60 is the specimen “B”. If it is determined that the measured value does not fall within the threshold range (NO in S03), the processor 201 determines that the hardness of the specimen 60 is "3" and that the specimen 60 is specimen "C".

Further, the processor 201 controls the gripping state of the robot hand 2 according to the hardness of the specimen 60 gripped by the robot hand 2. The processor 201 controls the gripping state of the robot hand 2 by controlling the amount of air (suction force) in the internal hollow portion 21 when the robot hand 2 is gripping the specimen 60 . For example, the processor 201 may control the force applied to the specimen 60 when the robot hand 2 grips the specimen 60 by controlling the drive amount of the fingers 2 by changing the suction force.

For example, a waveform corresponding to the hardness of the specimen 60 and control information corresponding to the hardness of the specimen 60 may be stored in the storage device 202 in association with each other. The control information is information for controlling the gripping state of the robot hand 2, and is, for example, information regarding the amount of drive of the fingers 2. The processor 201 may refer to the storage device 202 based on the waveform acquired from the charging sensor 41, acquire control information from the storage device 202, and control the grip state of the robot hand 2. Further, the processor 201 may control the movement (position) and tilt of the robot hand 2 based on the waveform acquired from the charging sensor 41 regarding the specimen 60.

Based on the waveform acquired from the charging sensor 41, the processor 201 can determine whether the specimen 60 gripped by the robot hand 2 is a good product or a defective product. In the storage device 202, data regarding a waveform depending on the hardness of the specimen 60 and data indicating whether the specimen 60 is a good product or a defective product are stored in association with each other.

The processor 201 compares data regarding the waveform according to the hardness and shape of the specimen 60 with the waveform acquired from the charging sensor 41, and determines whether the specimen 60 gripped by the robot hand 2 is a good product or a defective product. You may.

As described above, the hardness of the specimen 60 can be determined from the intensity (peak) of the response signal. For example, if the specimen 60 is an apple, it will be hard if the freshness is good (good product), and it will become soft (defective product) if the freshness deteriorates (rotten). Therefore, by adopting the following configuration, it is possible to determine whether the sample (apple) is good or bad.

Using the charge sensor 41, the strength of the response signal when the hardness of a defective product is investigated (intensity high, signal strength Ba) and the strength of the response signal when the hardness of a non-defective product is investigated (intensity (signal intensity Ca) is measured.

A value obtained by dividing the signal strength Ca by the signal strength Ba is set as a threshold value (TH1) for determining the quality of the specimen, and the processing of steps S01 and S02 in FIG. 9 is performed on the specimen 60. In the process of S02, if the measured value (signal intensity) is larger than 0 and smaller than the threshold value TH1, the sample is classified into the non-defective class (determined to be non-defective). Otherwise, the specimen is classified into a class of defective product (determined to be defective).

Thereby, for example, the processor 201 can perform fruit sorting for the sample 60 that is a fruit. Furthermore, for example, it is possible to separate fruits of poor quality into juices and to separate fruits with good quality into foods for raw consumption. Furthermore, the processor 201 uses AI (Artificial
The damage state of a specific fruit may be determined based on the waveform acquired from the charging sensor 41 in combination with Intelligent Intelligence) or deep learning.

Note that each of the plurality of fingers 2 may be made replaceable, or the fingers 2A and 2B and the actuator 3 may be made replaceable. In the case of a drive mechanism based on deformation tension of a dielectric elastomer, thread tension, or air pressure, it is easy to replace the fingers 2A and 2B and the actuator 3. The robot hand 2 may be attached to the arm portion 4 in a replaceable manner. Furthermore, by providing an imaging device such as a camera on the manipulator 207, the specimen 60 may be imaged by the imaging device, and the processor 201 may identify the type of the specimen 60.

As described above, the processor 201 controls the gripping state of the robot hand 2 based on the signal applied to the charging sensor 41 according to the hollow space. Thereby, the specimen 60 can be appropriately gripped. For example, the processor 201 can appropriately grip the specimen 60 by controlling the gripping state of the robot hand 2 depending on the hardness of the specimen 60.

(Processing example 2)
The response signal of the charging sensor 41, that is, the information (measured value) indicating the change in capacitance and the temporal change in current or voltage due to the change in pressure, is input to the determination device 200 and stored in the storage device 202. be done. The storage device 202 stores determination information regarding a plurality of types of specimens handled as the specimen 60 of the robot hand 2. The determination information may be acquired by reading determination information stored in advance in a storage device, read from an external storage device connected to the determination device 200, or acquired through timely communication.

The determination information may include one or more threshold values regarding the amount of change in level (size of level fluctuation) accompanying the supply and release of pressure. By comparing this threshold value with the measured value of the amount of change, it is possible to determine that a specimen larger than the threshold value and a specimen smaller than the threshold value are different types. For example, in a comparison between graph A and graph B in FIG. 6, large objects rise and reach a peak earlier than small objects. Therefore, for example, if a threshold value is set for the arrival time to the peak, the size of the specimen 60 can be determined.

Further, the determination information may include one or more threshold values regarding the decay time of the response signal (the time from when pressure supply is released (stopped) until the value of the response signal becomes 0). The type of specimen can be determined by comparing such a threshold value with the measured value of the decay time.

The type of specimen can also be determined based on whether at least one of an increase in the level during the pressure increase period and a decrease in the level during the pressure decay period is observed in the response signal (numeric value for determination).

In this way, the determination device 200 can identify the specimen using the measured value of the response signal regarding the specimen and the determination information. In addition, relative physical properties (hardness and elasticity) can be measured with respect to a comparative object. That is, the relative physical properties of the gripped specimen can be measured by comparing different types of objects or comparing objects of the same type with different shapes. Furthermore, even if there is no comparison target, the hardness and elasticity (returning force) of the specimen can be expressed in some unit. Furthermore, by determining the presence or absence of a phase inversion point in the signal, it is also possible to determine the shape and contact state of the gripped specimen.

FIG. 10 is a flowchart illustrating an example of sample determination processing performed by the processor 201. The process shown in FIG. 10 is a process for identifying (identifying) specimens a to e of different types (physical properties and shapes). In step S101, the processor 201 obtains a measured value of a response signal regarding the specimen. The measured value may be supplied from the electrometer 208 in real time, or may be read from the storage device 202 at a predetermined timing.

In step S102, the processor 201 analyzes the waveform to determine, with respect to the response signal (voltage change) corresponding to one pressure supply and its release, the phase change in the response signal period corresponding to the pressure increase period or the pressure decay period. Determine whether there is an inversion. If it is determined that there is a phase inversion (YES in S102), it is determined that the specimen to be identified is specimen c. On the other hand, if it is determined that there is no phase inversion (NO in S102), the process proceeds to step S103.

In step S103, the processor 201 analyzes the waveform to determine whether or not the decay time of the response signal corresponding to the pressure decay period is longer than the threshold value, regarding the response signal (voltage change) corresponding to one pressure supply and its release. Determine. If the decay time is longer than the threshold (there is a delay time), the process proceeds to step S104, and if it is determined otherwise, the process proceeds to step S105.

In step S104, the processor 201 determines whether the length of the delay time is equal to or greater than a threshold value. If the length of the delay time is determined to be greater than or equal to the threshold (YES in S104), the sample is sample e.
If not (NO in S104), the sample is determined to be sample d.

In step S105, the processor 201 determines whether the level (increase amount) of the response signal is larger than a threshold value with respect to the response signal (voltage change) corresponding to one pressure supply and release thereof, by analyzing the waveform. . If it is determined that the level is greater than the threshold (YES in S105), it is determined that the sample 60 is the sample b. On the other hand, if the level is below the threshold (NO in S105), the sample is determined to be sample a. In this way, it is possible to determine which of the samples a to e the sample 60 that the hollow charge sensor 40 has contacted is.

Note that the processor 201 is an example of a "judgment circuit", and the processing performed by the processor 201 as a determination circuit is performed by a processor other than a CPU, such as a plurality of CPUs, a CPU having multiple cores, a DSP, or a GPU. You can. Further, the processing as the determination circuit may be performed by an integrated circuit such as an ASIC or an FPGA, or a combination of a processor and an integrated circuit (such as a System-on-a-Chip (SoC)).

According to the above-described charging sensor 41 and the robot hand 2 to which the charging sensor 41 is applied, it is possible to appropriately grip an object (specimen) that easily deforms, such as a flexible container, and the robot hand 2 can properly grip an easily deformed object (specimen) such as a flexible container. It is possible to avoid spilling of contents (food, drink, etc.) due to deformation. Further, according to the determination device 200, the physical properties of the specimen 60, the state of the specimen 60 with respect to the hollow portion 44, the size of the specimen 60, etc. can be appropriately determined using the robot hand 2 to which the charging sensor 41 is applied. can. The configurations described in the embodiments described above can be combined as appropriate without departing from the purpose of the invention.

2...Robot hand (gripping part)
2A...Finger 2B...Finger (hand type dielectric)
10...Sensor body part 20...Counter electrode part 21...Inner hollow part 22...Air suction port 44 for sucking air inside the hollow part...Hollow part 45...In the hollow part Inside 46...Air suction port 41...Charging sensor 43...Output portions 43A, 43B...Lead wires 46...Air openings 51A, 51B...Charging part (charged body)
52B...Dielectric 52A...Dielectric 53A, 53B...Electrode 531B...Electrode 532B...Electrode 532C...Switch 54A, 54B...Support 55A...High resistance layer ( noise removal layer)
58... Contact surface 59... Pressing surface 60... Specimen 200... Judgment circuit 201... Processor 202... Storage device

Claims (11)

a first charging section including a first electrode and a first dielectric;
a second charging section including a second electrode paired with the first electrode and a second dielectric;
an extraction unit that extracts a signal generated by a change in capacitance between the first charging unit and the second charging unit from the first electrode,
The first dielectric body is formed using an elastic member and includes a hollow portion that is hollow inside;
The hollow part includes an opening through which air enters and exits between the inside and outside as the hollow part deforms due to contact with the specimen,
The extraction section is a charging sensor that extracts a signal generated due to the deformation of the hollow from the first electrode. The first charging section includes the first dielectric, the first electrode, a noise removal layer that removes noise in the signal, and a third electrode that is grounded together with the second electrode. The charge sensor according to claim 1, having a laminated structure. The charge sensor according to claim 2, wherein the noise removal layer has a volume resistivity of 10 ⁵ Ω or more. The charge sensor according to claim 2 or 3, further comprising a switch that grounds the first electrode. The charge sensor according to any one of claims 1 to 4, wherein the hollow portion is formed using at least one selected from silicone rubber, urethane rubber, acrylic rubber, natural rubber, and Teflon rubber. A robot hand including a first finger and a second finger different from the first finger,
The first finger is provided with the first charging portion according to any one of claims 1 to 5, and the second finger is provided with the second charging portion according to any one of claims 1 to 5. A robot hand characterized by being provided with a charging section. The robot hand according to claim 6, wherein the second finger is used as the second dielectric. Based on the shape of the signal taken out from the taking out part depending on the state of the specimen held by the robot hand according to claim 6 or 7 and the hollow part that is deformable by contacting the specimen, including a determination circuit that determines at least one of the physical properties of the specimen and the state of the specimen with respect to the hollow portion;
Sample determination device. 9. The specimen determining device according to claim 8, wherein the determining circuit determines the hardness of the specimen based on the magnitude of the level of the signal. The determination of the specimen according to claim 8 or 9, wherein the determination circuit determines the elasticity of the specimen based on the decay time of the level of the signal accompanying release of pressure supply from the specimen to the hollow part. Device. The determination of a specimen according to any one of claims 8 to 10, wherein the determination circuit determines the size of the specimen based on the timing of a rise of the signal accompanying contact of the specimen with the hollow portion. Device.

Images