Sensor, sensor system and method of sensing

Abstract

Disclosed herein a sensor, a sensor system and a method of sensing. The sensor, comprises: a multilayer base defining a thickness axis, the multilayer base comprising: a substrate layer; a sensing layer coupled to the substrate layer, the substrate layer being more rigid than the sensing layer; at least one Fiber Bragg Grating (FBG) embedded in the sensing layer, wherein the substrate layer and the sensing layer are collectively bendable relative to a neutral plane, wherein the at least one FBG is spaced apart from the neutral plane along the thickness axis, and a plurality of projections coupled to a sensing surface of the sensing layer and adjacent to the at least one FBG, wherein the plurality of projections defines at least one contact region on the sensing surface.

Description

SENSOR, SENSOR SYSTEM AND METHOD OF SENSINGCROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims the benefit of priority to the Singapore application no. 10202403432R filed 5 November, 2024, the contents of which are hereby incorporated by reference in their entirety for all purposes.TECHNICAL FIELD

[0002] This application relates generally to the field of sensing, and more particularly, to a sensor, a sensor system and a method of sensing.BACKGROUND

[0003] Integrating sensors of different natures for sensing and processing multiple stimuli presents a significant challenge to conventional sensing systems. This is particularly relevant for robotic systems, which rely on multiple sensor inputs. Robotics systems often rely on multiple tactile sensors positioned on manipulators or end-effectors for performing various tasks, such as repetitive pick-and-place tasks. However, the integration of multiple sensors, such as sensors for reflecting tool-object interaction forces and detecting slippage, to enable precise and reliable manipulation, is often challenging.SUMMARY

[0004] According to an aspect, disclosed herein a sensor. The sensor comprises a multilayer base defining a thickness axis, the multilayer base comprising: a substrate layer; a sensing layer coupled to the substrate layer, the substrate layer being more rigid than the sensing layer; at least one Fiber Bragg Grating (FBG) embedded in the sensing layer, wherein the substrate layerand the sensing layer are collectively bendable relative to a neutral plane, wherein the at least one FBG is spaced apart from the neutral plane along the thickness axis, and a plurality of projections coupled to a sensing surface of the sensing layer and adjacent to the at least one FBG, wherein the plurality of projections defines at least one contact region on the sensing surface.

[0005] According to another aspect, disclosed herein a sensor system. The sensor system comprises at least one sensor as described above; and an optical interrogator in optical communication with each of the at least one sensor, the optical interrogator being configured to receive an optical signal from each of the at least one sensor.

[0006] According to another aspect, disclosed herein a method of sensing. The method of sensing comprises: a sensor as described above; receiving a shear force along a shear sensing axis on the at least one contact region of the sensor, such that the at least one FBG undergoes tension to shift a FBG wavelength of an optical signal along a first shift direction; and receiving a normal force along a normal sensing axis on the at least one contact region of the sensor, such that the at least one FBG undergoes compression to shift the FBG wavelength of the optical signal along a second shift direction, wherein the shear sensing axis is transverse to the normal sensing axis, and wherein the second shift direction is opposing the first shift direction.BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Various embodiments of the present disclosure are described below with reference to the following drawings:FIG. 1 is a schematic diagram showing a robotic manipulator and a sensor system according to embodiments of the present disclosure.FIGs. 2 and 3 are parametric views of a robotic gripper with sensors attached according to embodiments of the present disclosure.FIGs. 4 to 6 are parametric view and partial parametric views of a prosthetic hand with sensors attached according to embodiments of the present disclosure.FIG. 7 is a parametric view of a sensor according to embodiments of the present disclosure.FIG. 8 is a sectional side view of the sensor of FIG. 7 along section A-A.FIGs. 9 to 1 IB illustrate the sensor under a shear force according to embodiments of the present disclosure.FIG. 12 illustrates a general working principle of a FBG.FIGs. 13 to 15 illustrate the sensor under a normal force according to embodiments of the present disclosure.FIGs. 16A to 16C are sectional views of a sensor according to various embodiments of the present disclosure.FIGs. 17A and 17B are top views of a sensor according to various embodiments of the present disclosure.FIGs. 18A and 18B are sectional views of a sensor according to various embodiments of the present disclosure.FIGs. 19A and 19B are parametric and side views of a sensor according to various embodiments of the present disclosure.FIG. 20 is a sectional view of a dual-sided sensor according to various embodiments of the present disclosure.FIG. 21 is a flowchart of a method of sensing according to various embodiments of the present disclosure.FIG. 22 is a schematic diagram illustrating a method of determining a neutral axis in a multilayer material.FIG. 23A shows a parametric, side, front, and top views of a sensor in a non-loaded state according to an exemplary implementation.FIG. 23B shows a parametric, side, front, and top views of the sensor of FIG. 23B in a loaded state (shear force applied).FIG. 23C is an image of a fabrication process of the proposed sensor.FIG. 23D are images of the sensor prototypes.FIG. 23E shows a parametric, side, front, and top views of a sensor in a non-loaded state according to an alternative implementation.FIG. 24A illustrates the working principle of FBG.FIG. 24B is a schematic diagram illustrating the working principle for contact normal force detection of the proposed sensor (due to compression).FIG. 24C is a schematic diagram to illustrating the working principle for slippage detection of the proposed sensor (due to elongation)FIG. 25 shows a parametric view of a test platform for contact force calibration.FIG. 26A is an image of a contact force test setup for a sensor with direct contact with a soft object.FIG. 26B shows the calibration and repeatability test result for the sensor with direct contact with the soft object.FIG. 26C shows the hysteresis test result for the sensor with direct contact with the soft object. FIG. 27A is an image of the contact force test setup for a sensor with direct contact with a rigid object.FIG. 27B shows the calibration and repeatability test result for the sensor with direct contact with the rigid object.FIG. 27C shows the hysteresis test result for the sensor with direct contact with the rigid object. FIG. 28A is an image of the contact force test setup for the sensor with direct contact with the rigid hollow chamber.FIG. 28B shows the calibration and hysteresis test results for the sensor with direct contact with the rigid chamber gap (15mm width).FIG. 28C shows the calibration and hysteresis test results for the sensor with direct contact with the rigid chamber gap (25mm width).FIG. 28D shows the trade-off between sensitivity and measurement range when the sensing unit contacts rigid material (with / without a gap).FIG. 28E shows the relationship between sensor sensitivity and the chamber gap width when the sensing unit contacts a rigid material (with / without a gap).FIG. 29 is a parametric view of a test platform for slippage tests.FIG. 30 shows the repeatability test result for slippage detections.FIG. 1 is an image of a demonstration in which a robotic gripper equipped with the sensor is used to open a threaded bottle fixed to the table.FIG. 32A shows the raw wavelength-shift signals from the two FBGs during two attempts to open the threaded bottle.FIGs. 32B and 32C show the wavelength shift signal converted into force for peak 3 of FIG. 32A.FIG. 33 is a schematic diagram of a processor system.DETAILED DESCRIPTION

[0008] The following detailed description is made with reference to the accompanying drawings, showing details and embodiments of the present disclosure for the purposes of illustration. Features that are described in the context of an embodiment may correspondingly be applicable to the same or similar features in the other embodiments, even if not explicitly described in these other embodiments. Additions and / or combinations and / or alternatives asdescribed for a feature in the context of an embodiment may correspondingly be applicable to the same or similar feature in the other embodiments.

[0009] In the context of various embodiments, the articles “a”, “an” and “the” as used with regard to a feature or element include a reference to one or more of the features or elements.

[0010] In the context of various embodiments, the term “about” or “approximately” as applied to a numeric value encompasses the exact value and a reasonable variance as generally understood in the relevant technical field, e.g., within 10% of the specified value.

[0011] As used herein, the term “and / or” includes any and all combinations of one or more of the associated listed items.

[0012] As used herein, terms “concurrently”, “simultaneously”, “at the same time”, or the like, may refer to events or actions that coincide or overlap within a period of time, regardless of whether the events start at the same time instant, and regardless of whether the events end at the same time instant.

[0013] As used herein, the term “multilayer base” may generally refer to a material base with varying rigidity along a thickness axis of the material. The “multilayer base” may include multiple layers of material with varying rigidity or modulus of elasticity, coupled to one another along the thickness axis. However, the term “multilayer base” need not necessarily refer to multiple physically discrete layers. Instead, the “multilayer base” may also include an integrally formed material with varying rigidity along a thickness axis of the material, such as a complex material or a composite material with multiple layers of rigidity.

[0014] As used herein, the term “neutral plane” may be a fictitious plane generally referring to an internal surface of a bent layer / plate / beam where there is no longitudinal stress or strain. The “neutral plane” may act as the boundary between the compressed (inner) portions / surfaces and tensioned (outer) portions / surfaces. The “neutral plane” comprises the neutral axis, which refers to the line where the neutral plane intersects the layer’ s / plate’s / beam's cross-section.

[0015] As used herein, the term “manipulator” may be used interchangeably with the terms “robotic manipulator”, “robotic arm”, “robot”, “end effector”, “prosthetic hand”, “prosthetic finger”, etc. and may generally refer to a programmable machine configured to perform one or more tasks by manipulating objects or tools. In some examples, the “manipulator” may comprise multiple links and joints. Typically, links are rigid connecting components that connect different sections of the manipulator, while their rigidity provides a solid structure. Joints are used to provide flexibility to the manipulator, allowing the manipulator to perform desired movements within a working space. The joints may move in different ways, and may facilitate exemplary motions such as linear, rotary, and revolutionary motions.

[0016] Integrating different sensors for sensing and processing multiple stimuli presents significant challenge to conventional sensing systems. This is particularly so for robotic systems, which rely on multiple sensor inputs and may be widely adopted across various sectors such as manufacturing, healthcare, logistics, food and beverage, and agriculture. For manipulator systems or robotic systems configured to perform tasks, such as repetitive pick- and-place tasks, the integration of multiple tactile sensors, such as sensors for reflecting toolobject interaction forces and detecting slippage, to enable precise and reliable manipulation is often challenging. Various industries may benefit from improved accuracy, efficiency, and safety in their operations by integrating this disclosed sensor into various grippers. The sensor offers the capability to process multiple stimuli with high repeatability, which provides the potential to facilitate more intelligent and adaptable automation solutions.

[0017] Conventional work on sensing and measuring tool-object interaction forces and detecting slippage are often met with limitations such as limited sensing ranges (e g. limited to a normal force sensing range of 0.03N), limited sensitivity and resolution, bulky and costly system due to integration of multi-modal sensors, susceptible to interference from externaldisturbances (such as magnetic sources) for magnetic based sensors, cross-sensitivity between sensors for optical based sensors, etc.

[0018] The present disclosure discloses a multifunctional sensor for force sensing and slippage detection. The proposed sensor may be an optical -based sensor integrating one or more Fiber Bragg Gratings (FBGs) within an optical fiber. The proposed sensor may be suitable for a manipulator, such as a robotic gripper or a prosthetic hand.

[0019] In various embodiments, the proposed sensor may meet the various performance matrices required for typical robotic systems, such as measurement range, sensing resolution, physical dimensions, rigidity, cross-sensitivity, compact size, easy of fabrication, cost, etc. Additionally, a machine learning model (or Al algorithm) may be employed to optimize signal processing, which improves sensor performance and enables multiple functionalities.

[0020] Departing from conventional approaches, the proposed sensor does not integrate multiple sensors, with each sensor detecting a single physical stimulus. The proposed sensor also does not integrate multiple modes of sensing for the detection of multiple physical stimuli.

[0021] According to various embodiments of the proposed sensor, synergies between the FBG and the unique structure of the substrate enable a common mode of sensing (via the FBG) for the detection of multiple stimuli (such as normal force and shear force). The proposed sensor may comprise a multilayer base or a multilayer structure with varying rigidity across a thickness axis. In various embodiments, the multilayer base may comprise a more rigid substrate layer and a less rigid sensing layer.

[0022] Tn various embodiments, the substrate layer and the sensing layer may be coupled, attached or integrated to one another. The substrate layer and the sensing layer may be collectively bendable, or in other words, the substrate layer and the sensing layer may collectively be bent into a bent state without separating from one another. The substrate layer and the sensing layer may define a neutral plane (and neutral axis) in the bent state. The opticalfiber with FBG may be embedded in the sensing layer spaced apart from the neutral axis along the thickness axis.

[0023] In avoidance of doubt, a person skilled in the art may understand that the neutral plane refers to a fictitious plane where there is no longitudinal stress or strain. The neutral plane comprises the neutral axis, which refers to the line where the neutral plane intersects the 1 ayer’s / pl ate ’s / b earn 's cross-section. As it may be appreciated, due to the varying rigidity in the multilayer base along the thickness axis, the neutral plane may not correspond or coincide with a centroid of the multilayer base.

[0024] In various embodiments, the proposed sensor may comprise an optical fiber embedded in the sensing layer. In an exemplary embodiment, the optical fiber may comprise at least one FBG provided or formed therein In other embodiments, the optical fiber may comprise multiple FBGs serially connected to one another.

[0025] In addition, due to the small size, the optical fiber with FBG is well-suited for embedding into a soft layer, such as the sensing layer. This also achieves a small overall size of the proposed sensor, allowing easy integration into the manipulator system, reducing the overall size and weight of the system. Multiple optical fibers with respective FBG(s) may also be connected to a single optical interrogator for simultaneously measuring or receiving optical signals from all the sensors concurrently.

[0026] Referring to FIG. 1, disclosed herein a sensor system 50 according to an aspect of the present disclosure. The sensor system 50 may be implemented for use on a manipulator 200 for tactile sensing. In various embodiments, the sensor system 50 may comprise at least one sensor 100, such as a pair of sensors 100 coupled to an opposing pair of gripping fingers 210 of the manipulator 200. Each of the pair of sensors 100 may comprise a multilayer base 110 with varying rigidity along a thickness axis of the multilayer base 110. In various embodiments, each of the multilayer base 110 may be less rigid than a respective attachment surface on thegripping fingers 210. For example, the multilayer base 110 may be formed from a polymeric material, and the attachment surface (or gripping fingers 210) may be made from a metallic material.

[0027] In addition, the multilayer base 110 may comprise multiple layers with varying rigidity along the thickness axis. In another example, the multilayer base 110 may include an integrally formed complex material with a continuously varying rigidity along the thickness axis 115. As such, the multilayer base 110 may comprise a “single layer” material with multiple rigidity along the thickness axis 115. It may be understood that the multilayer base 110 may correspond to a single layer or multiple layers, wherein the multilayer base 110 is characterised by varying rigidity along the thickness axis 115.

[0028] In various embodiments, a FBG 150 may be embedded in at least one layer of the multilayer base 110. The FBG 150 may be embedded in a relatively less rigid layer(s) of the multilayer base 110. The FBG 150 may be embedded in the multilayer base 110 via an optical fiber. The FBG 150 may be a uniform FBG, which offers robust, broadly applicable performance. In other embodiments, the FBG 150 may also include other FBG variants, such as tilted FBG, chirped FBG, phase-shifted FBG, long-period FBG, as may be used where advantageous without departing from the scope of the present disclosure. Further, in alternative embodiments, other types of strain gauges (such as resistive strain gauges) may be used in place of the FBG 150 to be embedded in the multilayer base 110.

[0029] In various embodiments, the sensor system 50 may further include an optical interrogator 300 in optical communication with each of the at least one sensor 100. In various embodiments, the optical interrogator 300 may be in optical communication with each of the FBGs 150. The optical interrogator 300 may be configured to receive an optical signal 152 from each of the at least one sensor 100 concurrently or sequentially.

[0030] In various embodiments, the sensor system 50 may further comprise a controller 900 or a processing system in signal communication with the optical interrogator 300. The controller 900 may be configured to implement a module to convert the optical signal 152 from each of the at least one sensor 100 into a respective sensing signal 154. The sensing signal 154 may correspond to a measure of a normal force (NF) or a measure of a shear force (SF). The measure of the normal force may correspond to a gripping force on an object 80, and the measure of the shear force may correspond to a slippage on the grip on the object 80.

[0031] In various embodiments, the controller 900 may be configured to convert the optical signal 152 from each of the at least one sensor 100 into the respective sensing signal 154 using a machine learning model. The machine learning model may comprise at least one or a combination of: a Multilayer Perceptrons (MLPs), a Convolutional Neural Networks (CNNs), a Long Short-Term Memory networks (LSTMs), a Gated Recurrent Units (GRUs), a Transformer-based architectures.

[0032] FIGs. 2 and 3 illustrate an exemplary implementation of the sensor system 50. The sensor system 50 may be implemented or used on a robotic gripper 200 (or the manipulator), according to various embodiments. The robotic gripper 200 may comprise a pair of gripper fingers 210. The pair of gripper fingers 210 may each be provided with a sensor 100. Each of the sensors 100 may be attached or coupled to a respective one of the pair of gripper fingers 210. Each of the sensors 100 may be in optical communication with an optical interrogator 300 such that optical signals from the sensors 100 may be transmitted and received by the optical interrogator 300.

[0033] FIGs. 4 to 6 illustrate another exemplary implementation of the sensor system 50. The sensor system may be implemented or used on a prosthetic hand 250 (or the manipulator), according to various embodiments. The prosthetic hand 250 may comprise a plurality of prosthetic fingers 260. The plurality of prosthetic fingers 260 may each be provided with asensor 100. Each of the sensors 100 may be in optical communication with an optical interrogator 300 such that optical signals from the sensors 100 may be transmitted and received by the optical interrogator 300. In some implementations, multiple sensors 100, each with at least one FBG 150 may be provided on a common optical fiber. For example, one or more prosthetic fingers 260 may be attached with two sensors 100, which are in optical communication via a common optical fiber.

[0034] In other implementations, the sensor system 50 may be configured for other use cases, including but not limited to: manufacturing and assembly, healthcare and health services, logistics and shipping, food and beverage, agriculture, etc.

[0035] FIGs. 7 and 8 illustrate a sensor 100 according to various embodiments of the disclosure. The sensor 100 comprises a multilayer base 1 10 defining a thickness axis 1 15. The multilayer base 110 may have a varying rigidity or modulus of elasticity along the thickness axis 115. In various embodiments, the multilayer base 110 may comprise a substrate layer 120 and a sensing layer 130 coupled to the substrate layer 120. The substrate layer 120 may be more rigid (or have a higher modulus of elasticity) than the sensing layer 130. The multilayer base 1 10 may further comprise at least one Fiber Bragg Grating (FBG) 150 embedded in the sensing layer 130. Each of the at least one FBG 150 may be formed in a respective optical fiber embedded in the sensing layer 130. The substrate layer 120 may be attached or coupled to the sensing layer 130 such that the substrate layer 120 and the sensing layer 130 are collectively bendable relative to a neutral plane 125. Due to the varying or different rigidity in the multilayer base 1 10 along the thickness axis 1 15, the neutral plane 125 may not correspond or coincide with a centroid of the multilayer base 110. The neutral plane 125 may be transverse to the thickness axis 115, such that the thickness axis 115 crosses the neutral plane 125. In some embodiments, the neutral plane 125 may be perpendicular or substantially perpendicular to the thickness axis 115.

[0036] In various embodiments, the at least one FBG 150 may be positioned spaced apart from the neutral plane 125 along the thickness axis 115. As such, the FBG 150 is located, disposed or positioned away from the neutral plane 125 and experiences longitudinal stress and / or strain arising from compression during bending. In some embodiments, the at least one FBG 150 may also be positioned spaced apart from a centroid (or a central plane) of the multilayer base 1 10.

[0037] Still referring to FIGs. 7 and 8, in various embodiments, a plurality of projections 140 may be coupled to a sensing surface 135 of the sensing layer 130. The plurality of projections 140 may extend away from the sensing surface 135 along the thickness axis 115. As shown in FIG. 7, the plurality of projections 140 may be formed as a plurality of ridges each extending along a first diametrical dimension (DI) of the multilayer base 1 10. In addition, the plurality of projections 140 may be formed as a plurality of ridges spaced apart from and parallel to one another along a second diametrical dimension (D2) ofthe multilayer base 110. In various embodiments, each of the plurality of projections 140 may be at least equally rigid or more rigid than the sensing layer 130. In various embodiments, the plurality of projections 140 may be integrally formed with the sensing layer 130, such as by use of additive manufacturing. In various embodiments, the plurality of projections 140 may be made of the same material as the sensing layer 130. Alternatively, the plurality of projections 140 and the sensing layer 130 may be independently formed, and coupled or attached together.

[0038] In various embodiments, the plurality of projections 140 may be adjacent to or neighbour to one or more of the at least one FBG 150 embedded in the sensing layer 1 0. The distance between the plurality of projections 140 and the at least one FBG 150 may be varied according to various considerations, such as sensitivity of the sensor 100, robustness to external disturbances, structural integrity ofthe sensor 100, etc.

[0039] In various embodiments, the plurality of projections 140 may define at least one contact region 105 on the sensing surface 135. Further, the sensor 100 may provide an optical signal (or sensing signal) responsive to a normal force (NF) and / or a shear force (SF) acting on the at least one contact region 105. The term “contact region” may refer to a region or surface of the sensor 100 contactable with an external object 80 for sensing, such as for tactile sensing. In addition, the term “contact region” corresponds to the provision of a single optical signal In an example where the sensor comprises a single FBG 150, one of the plurality of projections 140 adjacent to the single FBG 150 may be defined as a contact region 105, corresponding to a single sensing region providing a single optical signal. In other examples where there are multiple FBGs 150, the plurality of projections 140 may define multiple contact regions 105 providing multiple optical signals. Hence, the at least one contact region 105 may correspond to sensing region(s) on the sensor 100.

[0040] In exemplary embodiments without being limiting, the substrate layer (more rigid layer) may be made from polymers, such as: Polyethylene Terephthalate Glycol (PETG), Polycarbonate (PC), Polyethylene Terephthalate (PET), Acrylonitrile Butadiene Styrene (ABS), Polylactic Acid (PLA), etc. The substrate layer may typically have a Young’s modulus (Ei) of around 1.5 to 4.0 GPa (or 1500 to 4000 MPa), and in some instances within a range of approximately 1.8 to 2.6 GPa (e.g. PETG range, 1800 to 2600 MPa). The sensing layer (less rigid layer) may be made from polymers, such as: Thermoplastic Polyurethane (TPU), silicone elastomer, etc. The sensing layer may typically have a Young’s modulus (Ez) of around 0.5 to 100 MPa, and in some instances within a range of 5 to 30 MPa. In some embodiments, the relative rigidity between the substrate layer and the sensing layer may be in an order of 15 to 8000. In some instances, the relative rigidity between the substrate layer and the sensing layer may be in an order of 60 to 520.

[0041] In exemplary embodiments without being limiting, the substrate layer may have a thickness of 0.4 to 2mm, and in some instances within a range of 0.4 to 1mm. The sensing layer may have a thickness of 0.6 to 3mm, and in some instances, within a range of 0.6 to 1.5 mm. The protrusions may have a height in a range of 2 to 5mm. As such, the multilayer base may be thicker or thinner than the protrusions.

[0042] FIGs 9 to 1 1B illustrate an operation of the sensor 100 when subjected to a shear force (SF) according to various embodiments of the disclosure. Referring to FIGs. 9 and 10, the plurality of projections 140 of the sensor 100 may define a shear sensing axis 160. The shear sensing axis 160 may be transverse to the thickness axis 115. In some embodiments, the shear sensing axis 160 may be transverse to the plurality of projections 140. The shear sensing axis 160 may be generally parallel to the sensing surface 135 of the sensing layer 130. It may be appreciated that the shear sensing axis 160 corresponds to the sensing direction for an external shear force (such as a slippage) on the plurality of projections 140.

[0043] Fig. 12 illustrates the working principle of a Fiber Bragg Grating (FBG) wherein the FBG is under either tension or compression. The FBG may comprise multiple gratings with grating gaps (A) formed therebetween When the FBG is not under load, i.e. not under tension nor compression, the FBG outputs an optical signal corresponding to a reflected spectrum of the optical beam. When the FBG is subjected to tension, the grating gaps (A) increase (A’), resulting in a shift of FBG wavelength of the reflected spectrum towards a first direction (e g. increase). Conversely, when the FBG is subjected to compression, the grating gaps (A) decrease (A”), resulting in a shift of FBG wavelength of the reflected spectrum towards a second direction (e.g. decrease), wherein the second direction is opposite to the first direction. Hence, by observing the shifting of the FBG wavelength, it may be determined whether the FBG is undergoing tension or compression. Alternatively, the FBG may output an optical signal corresponding to a transmitted spectrum of the optical beam.

[0044] Further referring to FIGs. 11 A, 11B and 12, responsive to the at least one contact region 105 receiving a shear force (SF) along the shear sensing axis 160, the shear force on the plurality of projections 140 may deform the sensing layer 130 such that the at least one FBG 150 undergoes tension. This causes the at least one FBG 150 to extend, resulting in a shift in the FBG wavelength along a first shift direction 156. In avoidance of doubt, the at least one contact region 105 receiving a shear force may also correspond to a part or a portion of the at least one contact region 105 receiving the shear force.

[0045] In various embodiments, responsive to the at least one contact region 105 receiving a shear force (SF) along the shear sensing axis 160, the at least one FBG 150 undergoes tension (see FIGs. 11B and 12) to shift a FBG wavelength 155 along a first shift direction 156. The shifting of the FBG wavelength 155 may correspond to a shift in a frequency peak of the FBG wavelength 155.

[0046] It may be noted that the plurality of protrusions 140 may act to increase or amplify the extension experienced by the FBG 150, thus increasing the sensitivity of the sensor in response to a shear force (slippage). In addition, the plurality of protrusions

[0047] FIGs. 13 to 15 illustrate an operation of the sensor 100 when subjected to a normal force (NF) according to various embodiments of the disclosure. Referring to FIGs. 13 and 10, the plurality of projections 140 of the sensor 100 may define a normal sensing axis 170. The normal sensing axis 170 may be parallel to the thickness axis 115. In some embodiments, the normal sensing axis 170 may be parallel to the plurality of projections 140. The normal sensing axis 170 may be generally transverse to the sensing surface 135 of the sensing layer 130. The normal sensing axis 170 may be transverse to the shear sensing axis 160. In some embodiments, the shear sensing axis 160 may be perpendicular or substantially perpendicular to the nonnal sensing axis 170. It may be appreciated that the normal sensing axis 170 corresponds to thesensing direction for an external normal force (such as a gripping force) on the plurality of projections 140.

[0048] Further referring to FIGs. 12 and 15, responsive to the at least one contact region 105 receiving a normal force (NF) along normal sensing axis 170, the normal force on the plurality of projections 140 may deform the multilayer base (comprising both the sensing layer 130 and the substrate layer 120) such that the at least one FBG 150 undergoes compression. This causes the at least one FBG 150 to compress, resulting in another shift in the FBG wavelength along a second shift direction 157. In avoidance of doubt, the at least one contact region 105 receiving a normal force may also correspond to a part or a portion of the at least one contact region 105 receiving the normal force (as shown in FIG. 14).

[0049] In various embodiments, responsive to the at least one contact region 105 receiving a normal force (NF) along the normal sensing axis 170, the at least one FBG 150 undergoes compression (see FIG. 15) to shift the FBG wavelength 155 along a second shift direction 157. The second shift direction 157 is opposite to or in opposing direction from the first shift direction 156 as shown in FIG. 12. Similarly, the shifting of the FBG wavelength 155 may correspond to a shift in a frequency peak of the FBG wavelength 155.

[0050] As such, by observing for the shift in the FBG wavelength 155 in the first shift direction 156 enables the sensor 100 to sense a shear force (SF) corresponding to a slipping force or slippage on an external object. Similarly, by observing for the shift in the FBG wavelength 155 in the second shift direction 157 enables the sensor 100 to sense a normal force (NF) corresponding to a gripping force or gripping state on the external object.

[0051] The sensor 100 may realise both normal contact force measurement and slippage detection. The direction of the FBG wavelength 155 shift allows the differentiation between force levels and slippage occurrences, and the change rate of the FBG wavelength 155 shift may be used to calculate the slippage speed. Positioned with an offset distance from the neutralplane (or neutral axis) of the multilayer base, the FBG 150 experiences compression when a force is applied, causing the wavelength shift in one direction (e.g. to the left). The magnitude of the contact force may be calculated using the FBG wavelength 155 shift. Conversely, when slippage occurs, the protrusions or ridges deform causing the sensing layer to stretch, thus resulting in tension / elongation of the FBG 150 and a corresponding wavelength shift in another direction (e.g. to the right). As such, the sensor 100 may be used for the detection of slippage and slippage speed.

[0052] In embodiments with multiple FBGs, each FBG may be configured in sensing each of the shear force (SF) and the normal force (NF) such that the gripping state and slippage may be concurrently sensed.

[0053] It may also be appreciated that the shift in the FBG wavelength 155 along the first shift direction 156 may be a positive shift or a negative shift along the frequency axis. As such, the shift in the FBG wavelength 155 along the first shift direction 156 may be to a corresponding negative shift or a corresponding position shift along the frequency axis.

[0054] In alternative embodiments, other types of strain gauges (such as resistive strain gauges) may be used in place of the FBGs 150. In these resistive strain gauges, the resistance decreases under compression, allowing for contact force measurement, and increases under tension during slippage.

[0055] FIGs. 16A to 16C illustrate various embodiments of the plurality of projections 140. Referring to FIG. 16A, in various embodiments, each of the plurality of projections 140 may be tilted relative to the thickness axis 1 15 towards a tilt direction 142 parallel to the shear sensing axis 160. The tilting of the plurality of projections 140 enables an amplification of the tension applied to the FBG 150 in response to the shear force (SF) in a direction opposing to the tilt direction 142. In various embodiments, each of the plurality of projections 140 may have a respective height (Hl) along the thickness axis 115 larger than a thickness (T) of themultilayer base 110 along the thickness axis 115. Further, each of the plurality of projections 140 may comprise a flat top surface 141 which enables a larger contact surface with an external object 80.

[0056] Referring to FIG. 16B, in various embodiments, the plurality of projections 140 may be configured as bumps, friction or texture elements. This enables the plurality of projections 140 to enhance the gripping effect by reducing potential slippage. In various embodiments, each of the plurality of projections 140 has a respective height (H2) along the thickness axis 115 smaller than a thickness (T) of the multilayer base 110 along the thickness axis 115.

[0057] Further referring to FIG. 16C, in various embodiments, the plurality of projections 140 may extend substantially parallel to the thickness axis 115. This allows the sensing of shear forces (SF) from both directions.

[0058] FIG. 17A illustrates a top view of a sensor 100 according to various embodiments of the disclosure. The sensor 100 may comprise a plurality of first projections 140A adjacent to a first FBG 150A, and a plurality of second projections 140B adjacent to a second FBG 150B. The first FBG 150A may be parallel to the second FBG 150B. Therefore, the plurality of first projections 140A of the sensor 100 may define a first contact region 105 A, and the plurality of second projections 140B may define a second contact region 105B, when the sensor 100 is in contact with an external object 80. Each of the plurality of first / second contact regions 105A / 105B may correspond to respective ones of a plurality of FBGs 150A / 150B. For example, the contact region 105 A may correspond to the first FBG 150A, and the contact region 105B may correspond to the second FBG 150B This enhances the sensing location accuracy of the sensor 100, providing a plurality of contact regions and sensing regions. The first FBG 150A and the second FBG 150B may be located, disposed or positioned at respective different planes along the thickness axis 115 offset from the neutral plane 125.

[0059] Fig. 17B illustrates a top view of another sensor 100 according to various embodiments. Similarly, the sensor 100 may comprise a plurality of first projections 140 A adjacent to a first FBG 150A, and a plurality of second projections 140B adjacent to a second FBG 150B. However, the first FBG 150A may be transverse to the second FBG 150B. In addition, the first FBG 150A may be is offset from the second FBG 150B along the thickness axis 1 15 The diagonal placement of the optical fiber and hence the FBGs 150A / 150B, enable not only slippage detection but also determination of slippage direction. The term “transverse to” may generally correspond to a first reference extending across a second reference, such as a first reference forming an angle with a second reference. In an example, the first FBG 150A may form a 30-degree angle with the second FBG 150B. In another example, the first FBG 150A may form a 45-degree angle with the second FBG 150B. In another example, the first FBG 150A may form a 60-degree angle with the second FBG 150B. In another example, the first FBG 150A may form a 90-degree angle with (and hence perpendicular to) the second FBG 150B.

[0060] Referring to FIG. 18A, in various embodiments, the plurality of first projections 140A may be tilted relative to the thickness axis 1 15 towards a first tilt direction 142. In addition, the plurality of second projections 140B may be tilted relative to the thickness axis 115 towards a second tilt direction 144, wherein the second tilt direction 144 is opposite to the first tilt direction 142. In various embodiments, the plurality of first projections 140A may form a first tilting angle relative to the thickness axis 115. The first tiling angle may be equal to a second tilting angle formed between the plurality of second projections MOB relative to the thickness axis 115. As such, the plurality of first projections 140A may be generally symmetrical with the plurality of second projections MOB relative to the thickness axis 115. This enables sensing sensitivity in response to an external shear force to be generally uniform in both tilt directions142 / 144.

[0061] Further referring to FIG. 18B, in various embodiments, the plurality of first projections 140A may form a first tilting angle relative to the thickness axis 115. The first tiling angle may be different from a second tilting angle formed between the plurality of second projections 140B relative to the thickness axis 115. As such, the plurality of first projections 140A may be asymmetrical with the plurality of second projections 140B relative to the thickness axis 115. This enables sensing sensitivity in one tilt direction (e g. first tilt direction 142) to be higher than in another tilt direction (e.g. second tilt direction 144) in response to an external shear force. This may be useful in instances where slippage is more likely to occur in one direction than the other, for example, in the gravity direction.

[0062] Referring to FIGs. 19A and 19B, in various embodiments, the substrate layer 120 may comprise a pair of supporting legs 122 coupled to opposing ends of the substrate layer 120. The pair of supporting legs 122 forming a chamber that allows more room for deformation when the sensor 100 is subjected to a normal force (NF) thus improving sensor sensitivity, such as compression force sensitivity. In addition, the plurality of protrusions 140 may be positioned adjacent to one side of the sensor to improve elongation of the FBG 150 during slippage.

[0063] FIG. 20 illustrates a dual-sensing surface sensor 100 according to various embodiments. The sensor 100 may comprise a multilayer base 110 defining a thickness axis 115. The multilayer base 110 may comprise a substrate layer 120 and a first sensing layer 130A coupled to the substrate layer 120. The multilayer base 110 may further comprise a plurality of Fiber Bragg Gratings (FBGs) 150A / 150B embedded in the first sensing layer 130A. A plurality of first projections 140A and a plurality of second projections 140B may be coupled to a first sensing surface 135A of the first sensing layer 130A. The plurality of first projections 140A and a plurality of second projections 140B may extend away from the first sensing surface 135A along the thickness axis 115. The plurality of first projections 140A may be adjacent toor neighbour to the first FBG 150A. The plurality of second projections 140B may be adjacent to or neighbour to the second FBG 150B.

[0064] The multilayer base 110 may further comprise a second sensing layer 130B coupled to the substrate layer 120. The first sensing layer 130A and the second sensing layer 130B may be coupled to opposing surfaces of the substrate layer 120. The multilayer base 110 may further comprise a plurality of Fiber Bragg Gratings (FBGs) 150C / 150D embedded in the second sensing layer 130B. A plurality of third projections 140C and a plurality of fourth projections 140D may be coupled to a second sensing surface 135B of the second sensing layer 130B. The plurality of third projections 140C and a plurality of fourth projections 140D may extend away from the second sensing surface 135B along the thickness axis 115. The plurality of third projections 140C may be adjacent to or neighbour to the third FBG 150C The plurality of fourth projections MOD may be adjacent to or neighbour to the fourth FBG 150D.

[0065] In various embodiments, the substrate layer 120 may be more rigid than each of the first sensing layer 130A and the second sensing layer 130B. The substrate layer 120 may be attached or coupled to the first sensing layer BOA and the second sensing layer 130B such that the substrate layer 120, the first sensing layer 130A and the second sensing layer 130B are collectively bendable relative to a neutral plane 125.

[0066] In various embodiments, the FBGs 150A / 150B / 150C / 150D may be positioned spaced apart from the neutral plane 125 along the thickness axis 115. As such, the FBGs 150A / 150B / 150C / 150D are disposed away from the neutral plane 125 and experiences longitudinal stress and / or strain during bending.

[0067] In various embodiments, the plurality of first projections 140A may define a first contact region 105A. The plurality of second projections MOB may define a second contact region 105B. The plurality of third projections 140C may define a third contact region 105C. The plurality of fourth projections MOD may define a fourth contact region 105D. 1

[0068] In various embodiments, on one side of the sensor 100, the plurality of first projections 140A may be relatively more sensitive to a shear force along the direction 144, and the plurality of second projections 140B may be relatively more sensitive to a shear force along the direction 142. Similarly, on another side of the sensor 100, the plurality of third projections 140C may be relatively more sensitive to a shear force along the direction 142, and the plurality of fourth projections 140D may be relatively more sensitive to a shear force along the direction 144.

[0069] According to another aspect of the disclosure, disclosed here is a method of sensing using the proposed sensor, according to various embodiments. FIG. 21 is a flowchart illustrating a method of sensing 700. The method of sensing 700 comprises: using the proposed sensor, in 710: receiving a shear force along a shear sensing axis on the at least one contact region of the sensor, such that the at least one FBG undergoes tension to shift a FBG wavelength of an optical signal along a first shift direction; and in 720: receiving a normal force along a normal sensing axis on the at least one contact region of the sensor, such that the at least one FBG undergoes compression to shift the FBG wavelength of the optical signal along a second shift direction. In various embodiments, the shear sensing axis is transverse to the normal sensing axis. In various embodiments, the second shift direction is opposing the first shift direction.

[0070] In various embodiments, the method 700 further comprises in 730: converting an optical signal from each of the at least one sensor into a respective sensing signal In various embodiments, the method 700 further comprises in 740: converting the optical signal from each of the at least one sensor into the respective sensing signal using a machine learning model. In various embodiments, the machine learning model comprises at least one of: a Multilayer Perceptrons (MLPs), a Convolutional Neural Networks (CNNs), a Long Short-Term Memory networks (LSTMs), a Gated Recurrent Units (GRUs), and Transformer-based architectures.

[0071] Each of these models may be capable of learning the nonlinear mapping between the FBG spectrum and physical parameters such as contact force, slippage force, velocity, or angle.By employing these algorithms, the system may reliably transform low-level wavelength shifts into high-level semantic outputs (e.g., force magnitude, slippage occurrence, and direction), which may be used to support real-time feedback and control in robotic manipulation, prosthetic systems, or medical devices. Importantly, the proposed framework is not restricted to a particular model or algorithm. Instead, the framework allows for the integration of any suitable Al model as the technology evolves, ensuring long-term adaptability and scalability.

[0072] Neutral plane or Neutral axis

[0073] The following section outlines an exemplary method (out of multiple methods) in determining a neutral axis (and hence neutral plane) of a multilayer structure (or multilayer base). The multilayer base may be treated as a two-layer composite material. The sensing layer (more elastic) has a Young’s modulus Ei and thickness hi, while the substrate layer (more rigid) has a Young’s modulus E2 and thickness I12, as illustrated in FIG. 22.

[0074] A coordinate system is defined with the origin of the Y-axis located at the neutral axis of the composite beam. Let h' denote the distance from the interface (between the lower and upper rock beams) to the neutral axis. Since Ei is typically on the order of MPa while E2 is on the order of GPa, the condition for the neutral axis of the composite beam to lie within the lower rock beam is fulfilled: Eihi2< E2I122, which indicates that the neutral axis shifts toward the bottom rigid layer.

[0075] Assuming both materials are within linear elastic ranges, based on static equilibrium principles, the resultant force on any cross-section of the beam must equal zero, hence:Thus, the neutral axis position may be derived as:

[0076] Exemplary Implementation

[0077] The present section describes an exemplary implementation of the proposed sensor. The proposed sensor comprises a rigid-elastic (Compliant-flexible) configuration The proposed sensor comprises three layers: i) a thin and relatively rigid substrate layer as the base, ii) a thin and elastic substrate layer at the middle with embedded FBG sensors, and iii) a top layer made of deformable structures / pattems such as inclined ridges. The combination of rigid and elastic materials enhances the sensor’s robustness while maintaining its flexibility and elasticity.

[0078] The optical sensor (FBG) or optical fiber may be fused within the elastic substrate layer during the 3D printing process, with the FBG located with an offset from the neutral axis of the base material. Embedding the sensor within the elastic substrate layer allows the normal contact force to be transmitted to the FBG sensor, allowing it to bend upwards and resulting in a negative wavelength shift due to compression. Additionally, the inclined protrusions or ridges play a crucial role in maintaining soft / gentle contact while amplifying the sliding effects, which results in a positive wavelength shift in the FBG due to elongation. The direction of the wavelength shift helps differentiate between varying force levels and slippage occurrences. The inclined protrusions allow for a gradual change in the contact surface, which increases the relative motion between the surface and the object in contact, thereby amplifying the slippage signal. At the same time, the soft / gentle contact is preserved, ensuring that the overall sensorobject interaction remains gentle and compliant.

[0079] Tn other implementations, a dual FBG array can be used optionally to detect and compensate for temperature effects. This sensor can be integrated with various end-effectors, including both soft and rigid grippers, to measure contact force and detect slippage across diverse industries, such as manufacturing, healthcare, logistics, food and beverage, and agriculture. It can also be employed in embodied intelligence systems, such as prosthetic limbs.Moreover, the sensing method is not restricted to FBG sensors, it can be extended to other types of resistive strain gauges. In these gauges, the resistance decreases under compression, allowing for contact force measurement, and increases under tension during slippage.

[0080] Also disclosed is a new method for embedding the FBG directly into a rigid-elastic substrate (or multilayer base) during the 3D printing process. Generally, FBGs are embedded into grooves or onto the surface of a rigid subtract using adhesives or integrated into a flexible material (such as silicone) through a curing process during molding or casting. However, integrating FBG into a 3D-printed substrate typically needs two steps: 1) pre-designing a channel or a cavity in the model and printing the model; 2) applying epoxy or another adhesive in a post-printing process to bond the FBG sensor within the 3D-printed substrate. The proposed method simplifies the process by eliminating the need for creating channels / cavities or applying adhesives, which enables seamless sensor integration directly during the 3D printing process.

[0081] Sensor configuration and fabrication

[0082] Referring to FIG.23A, the exemplary sensor or sensing unit 1100 comprises a multilayer structure. As shown in the side view, the bottom layer 1111 is made of rigid material such as PETG The middle layer 1 1 12 is made of elastic material such as TPU An optical fibre with FBG 1113 is embedded within the elastic substrate material. The top layer 1114, comprises multiple inclined ridges (or protrusions) 1115a~l 115j, is made of elastic material such as TPU. This configuration allows normal contact force to be transmitted to the FBG sensor, causing it to bend upwards and create compression on FBG.

[0083] FIG. 23B illustrates a scenario where an external force causes slippage or sliding on the top layer 1124 of the detection device 1120, with the force directed against the inclined ridges. The bottom layer 1121 is secured to a gripper or a fixture. When the inclined ridges 1125a ~1125j, which initially slope to the left, start bending towards the right, they create ashear effect on the middle layer 1122. This deformation can cause the optical fibre with FBG 1123 in the middle layer to elongate.

[0084] The proposed sensing unit was fabricated in a single process using the multi -material FDM 3D printer Original PrusaXL. Two independent tool heads were used, one loaded with rigid filament PETG whereas the other loaded with elastic filament TPU. The layer height for each print is set to 0.2 mm. Both the base rigid substrate layer and the middle elastic substrate layer have a height of 0.6 mm. After completing the printing of the base layer with PETG, the first two 3D printing layers of the elastic material are printed using TPU. The process is then paused to allow the optical fibre with polyimide coating to be positioned on top of the TPU layer. Pretension is applied to the optical fibre to prevent slacking by using the high-temperature Kapton tape, as depicted in FIG. 23C. Subsequently, the printing process resumes and the heat from the printing process helps fuse the TPU material around the optical fibre. The TPU is printed at a temperature between 220°C and 250°C. The polyimide-coated fibers can withstand temperatures up to 350°C for a short time and 300°C for continuous use. Photographs of the sensing unit prototypes are presented in Fig. 23D.

[0085] The 3D printer described in this invention is not limited to any specific model as disclosed. It may be known that any 3D printing device with similar capabilities may be used to achieve the purposes of this present disclosure. This section describes the workflow specific to a particular model of 3D printer. While the workflow may differ when using other printers, it does not affect the functionality of the invention. Additionally, the material is not limited to PETG and TPU and may be any other material with a comparable or suitable range of Young's modulus.

[0086] In alternative embodiments as illustrated in Fig. 23E, the sensor 1130 comprises a rigid substrate 1131 with supporting legs, forming a chamber that improves compression force sensitivity. In addition, the inclined ridges 1135a~l 135g may be positioned adjacent to one sideof the grating area of the optical fiber 1133 to ensure elongation occurs when slippage takes place. Other alternative designs, such as diagonal placement of the sensing fiber combined with multiple FBGs, as well as variations in ridge orientation, enable not only slippage detection but also determination of slippage direction. Additionally, the sensor may be treated as a single sensing unit, with its measurement degrees of freedom (DOF) extended by positioning FBG sensors and inclined ridges at different angles to detect slippage in multiple directions.

[0087] In the following section, described herein, a single-core optical fiber with Fiber Bragg Grating (FBG), recoated with polyimide coating, as an exemplary implementation. As shown in FIG. 24A, the single-core optical fiber 1200 comprises a coaxial core 1201, and a cladding 1202 covered by a coating 1203. A small amount of light is reflected at each grating 1204. The wavelength at which this reflection occurs is called the Bragg wavelength and the condition is called the Bragg condition:AD= 2 neA, (3) is the reflected wavelength or Bragg wavelength, neis the effective refractive index, and A is the grating period. If the light with wavelengths does not satisfy the Bragg condition, they will be transmitted through the fiber and exit the end of fiber.

[0088] The shift in the Bragg wavelength depends on the effective refractive index and the grating period which are a function involving both strain and temperature, as in:where kr, AT, kE, and s are the thermal sensing coefficient, temperature change, strain sensing coefficient of the fiber material, and axial strain along the fiber, respectively.

[0089] A dual FBG array may be used as an additional strain-free FBG serves as a temperature sensor, enabling the primary FBG to be temperature compensated. When the sensor is applied at constant laboratory room temperature, the temperature cross-sensitivity is minimized, and the wavelength shift ABbecomes proportional to strain change, as in:

[0090] As shown in FIG. 24A, when the FBG experiences tension, the Bragg wavelength shifts to the right on the reflected spectrum, indicating a positive wavelength shift. When the FBG experiences compression, the Bragg wavelength moves to the left on the spectrum, indicating a negative wavelength shift.

[0091] FIG. 24B illustrates scenario 1210, where a contact force F is applied by object 1212 to the proposed sensing unit 1211, which can be attached to either an elastic or rigid gripper. The FBG fiber 1215 is located with an offset 1216 from the neutral axis of the base material, which comprises a base rigid substrate layer 1213 and an elastic substrate layer 1214. When the base material bends upwards due to the normal contact force exerted by the object, the FBG sensor experiences compression, which leads to a negative wavelength shift.

[0092] FIG. 24C illustrates scenario 1220, where an object 1221 slides / slips against the inclined ridges 1222a~ 1222g of the proposed sensing unit 1223. As shown in the zoom-in views, these inclined ridges amplify the sliding effect against the inclined direction, making the thin elastic layer 1225 stretch and creating tension on the FBG 1224. When the object slides against ridges closer to the free end of the optical fiber, such as 1222d ~ 1222g, the stretching effect becomes more pronounced, which results in a positive wavelength shift.

[0093] Notably, the proposed sensor may be treated as a modular sensing unit, and by incorporating multiple layers or configurations, it can be extended into multi-DOF slippage sensors or adapted for various other tactile sensing applications.

[0094] Additionally, artificial intelligence (Al) algorithms, including machine learning, may be employed to enhance data analysis and improve the sensor's performance in complex environments. The raw wavelength shift spectrum collected from the sensor may first undergo preprocessing steps such as feature extraction and filtering. The processed data may then be fed into various machine learning or deep learning models, including but not limited to MultilayerPerceptrons (MLPs), Convolutional Neural Networks (CNNs), Long Short-Term Memory networks (LSTMs), Gated Recurrent Units (GRUs), Transformer-based architectures, etc.

[0095] Experiments

[0096] Normal contact force test platform and test results

[0097] FIG. 25 shows a test platform 1300 for determining the normal contact force and the wavelength shift of the sensing unit. As shown in FIG. 25, an L-shaped fixture 1302 was secured to a perforated plate 1301. An adaptor 1303 was attached to the fixture using screws, serving as a housing of a linear actuator 1304. The L-shape fixture 1302 had a slot that allowed the height of the linear actuator 1304 to be adjusted by loosening and retightening the screws as needed. Directly beneath the linear actuator 1304, the proposed sensor or sensing unit 1305 was placed on top of a donut-shaped load cell 1307 within housing 1306 that was connected to fixture 1302. For the wiring connections, the FBG fiber of the sensing unit 1305 was connected to an interrogator 1308 (Micron Optics si-255), which automatically captures the real-time wavelength data. The linear actuator 1304 was connected to a controller 1309 (Raspberry Pi 4 Model B). The load cell 1307 was connected to a DAQ (Data Acquisition) board 1310 for force data collection. In this normal contact force test, the linear actuator 1304 was commanded to move downwards and upwards to provide loading and unloading towards the sensing unit 1305. When the linear actuator touches the top layer of the sensing unit 1305 and continues to move downward, the middle elastic layer deforms and bends. Due to the offset position of the optical fibre at the base, this bending results in compression of the fibre, causing a decrement in the wavelength, indicated by a wavelength shift to the left on the reflection spectrum.

[0098] To study the sensitivity of the sensing unit, three categories of normal contact force tests were carried out based on the different objects it contacted: soft, rigid, and rigid with a gap. FIG. 26A shows a photograph taken during the test when the base rigid substrate layer of the sensing unit is in contact with a soft object. This soft object is 3D-printed using urethaneacrylate resin. To evaluate the repeatability of the sensing unit, three runs of tests were performed.

[0099] As shown in FIG. 26B, the results demonstrated good repeatability, with the outcomes of the three tests nearly overlapping. A second-order polynomial equation, y = 62.469 x2— 1223.833 x — 266.625, was curve-fitted as the calibration equation, yielding an R-square value of 0.995, where y represents the wavelength shift (pm) and x is the contact force (N). The total wavelength shift is approximately -6232.5 pm, with a corresponding force measurement range of about 8.25 N. The sensitivity is calculated as approximately - 755.45 pm / N, determined by dividing the wavelength shift by the measurement range (-6232.5 pm / 8.25 N). Additionally, another test was run to examine the hysteresis of the sensing unit. The results, shown in FIG. 26C, indicate no significant hysteresis was observed.

[0100] Fig. 27A is a photograph captured during the test when the bottom layer of the sensing unit is directly touched with a rigid object (metal of the load cell). Similarly, to investigate the repeatability of the sensing unit, three runs of tests were performed. As shown in FIG. 27B, the results demonstrated good repeatability, with the outcomes of the three tests nearly overlapping. A second-order polynomial equation, y — 2.115 x2— 135.643 x — 114.547, was curve-fitted as the calibration equation, with an R-square value of 0.990. The total wavelength shift is approximately -2421.6 pm, with a corresponding force measurement range of about 29.7 N. The sensitivity is calculated as approximately -81.53 pm / N, determined by dividing the wavelength shift by the measurement range (-2421.6 pm / 29.7 N). Also, another test was run to examine the hysteresis of the sensing unit. The results, shown in FIG. 27C, indicate no significant hysteresis was observed. It is worth noting that the sensitivity of the sensing unit is much higher when in contact with soft material compared to rigid material.

[0101] In addition, tests were conducted with the base rigid substrate layer of the sensing unit in contact with two other rigid materials, each featuring a chamber gap - one with a widthof 15mm while the other with 25mm, as shown in FIG. 28A. Calibration and hysteresis tests were carried out to assess sensor’s performance in these scenarios. As shown in FIG. 28B, when in contact with rigid material having a 15mm -width chamber gap, the sensor exhibited good linearity between the wavelength shift and contact force, with a linear fitting equation of y = —1462.128 x — 532.168, and an R-square value of 0.963. However, some degree ofhysteresis was observed in this scenario. Similarly, Fig. 28C illustrates the sensor’s performance when in contact with a rigid material having a 25mm-width chamber gap. Once again, good linearity between the wavelength shift and contact force was observed, with a linear fitting equation of y = —4779.199 % — 78.227 and an R-square value of 0.942. Nevertheless, the hysteresis effect becomes more obvious in this scenario.

[0102] Table below summarizes the sensitivity and measurement range of the sensing unit when in contact with different materials. It is important to note that there is a trade-off between the sensitivity and measurement range. As the measurement range increases, the sensitivity decreases. FIG. 28D highlights this trade-off between sensitivity and measurement range when the sensing unit contacts various rigid structures (test groups II, III-A, and III-B). The fitting exponential equation is y = 2627x1 015, with an R-Square of 0.9975. FIG. 28E further reveals the relationship between sensor sensitivity and the chamber gap width when the sensing unit contacts various rigid structures (test groups II, III-A, and III-B). A second-order polynomial function, y = 9.5867x2— 51.76x + 81.53, is fitted with an R-square value of 1. The two figures provided potential guidance for configuring the bottom layer structure of the sensing unit, based on the desired measurement range and sensitivity.Table 1 . Sensitivity and measurement range of the sensing unit when in contact with different materials

[0103] Slippage detection platform and test results

[0104] Referring to FIG. 29, a test platform 1700 was built to study the wavelength shifting behaviours when there is slippage on the top layer of the sensing unit. As shown in FIG. 29, a 3-axis XYZ translation stage 1702 was secured to a perforated plate 1701. An adaptor 1703, serving as a housing of a linear actuator 1704, was attached to a plate on the 3-axis XYZ translation stage 1702. A cap 1705 with a semi-circular feature was affixed to the tip of the linear actuator 1704. The proposed sensing unit 1706 was positioned directly beneath the linear actuator 1704, on top of solid connector 1707 mounted on a donut-shaped load cell 1708 within housing 1709 that was connected to the perforated plate 1701. For the wiring connections, the FBG fibre of the sensing unit 1706 was connected to an interrogator 1710 (Micron Optics si- 255), which automatically captures the real-time wavelength data. The linear actuator 1704 was connected to a controller 1711 (Raspberry Pi 4 Model B). The load cell 1708 was connected to a DAQ (Data Acquisition) board 1712 for force data collection. The home position was defined when cap 1705 was aligned right above the grating area of the sensing unit 1706. The steps for the slippage detection test were as follows:(1) Adjusted the z-axis knob of XYZ translation stage 1702 until the semi-circular feature of the cap 1705 touched the top layer of the sensing unit 1706 and applied a certain level of compression force.(2) The linear actuator 1704 was moved backward and created a slippage effect against the inclined ridges on the sensing unit 1706 until the cap 1705 fully disengaged from it.(3) Tuned the z-axis knob of XYZ translation stage 1702 upward to lift the cap 1705, and adjusted the y-axis knob to move the cap 1705 forward until it returned to its home position.The steps (1) to (3) were repeated three times.

[0105] FIG. 30 presents the corresponding test results, which may be divided into three distinct regions: load (1801a ~ 1801c), slip (1802a ~ 1802c), and release (1803a ~ 1803c). During the loading phase, the FBG sensor underwent a compression effect as mentioned in the previous subsection, resulting in a negative wavelength shift. In the slippage phase, the FBG sensor experienced elongation due to the shear force, causing the wavelength to shift to the right and become positive. When the semi-circular feature of the cap 1705 was gradually released from the sensing unit 1706, the wavelength shifts back to its initial value, causing the wavelength shift to gradually return to zero. This consistent behavioural pattern occurs across all three repetitions not only demonstrates the sensor’s good repeatability in detecting slippages but also confirms its ability to measure normal contact force

[0106] Examplary Applications

[0107] As described previously, the proposed sensor may be integrated either at the fingertip of a prosthetic finger (see FIGs. 4 to 6) or at the tool tip of a robotic gripper (see FIGs. 2 and 3). In this configuration, multiple FBGs may be inscribed along the optical fiber. Grating areas may be embedded within the middle soft substrate layer to measure contact force and detect slippage, while a separate strain-isolated grating region may be dedicated to temperature sensing.

[0108] As illustrated in FIG. 31, a demonstration was conducted where a robotic gripper equipped with the proposed multi-FBG sensor attempted to open a threaded bottle fixed to the table. The sensor experienced compression when the gripper maintained stable contact with the bottle lid, which led to a negative wavelength shift. During the first unscrewing attempt, slippage occurred between the gripper and the lid, causing partial wavelength recovery and introducing noise into the signals on both FBGs. When the gripper touches the bottle lid, the sensing structure bends upward, which makes the embedded FBG in compression and causesnegative wavelength shift. When slip occurs, engagement between the bottle lid’ s line-patterned surface and the sensor ridges locally elongates the FBG, which leads to short positive wavelength shifts superimposed on the negative baseline. In the second attempt, however, the gripper achieved secure contact and opened the bottle without slippage. As a result, the sensing signals remained clean, exhibiting only a negative wavelength shift corresponding to the compressive force. The raw signal results are shown in FIG. 32A. FTGs. 32B and 32C show the wavelength shift signal converted into force for peak 3 of FIG. 32A. The results show that a single off-neutral-axis FBG realizes reliable force-slip differentiation from the same optical signal. For practical applications, signal-processing methods — such as applying filters — may be employed to suppress undesired signal components.

[0109] The proposed sensor system and sensor may be applicable or implementable across various applications or industries by providing the tool -object contact status, including contact force measurement and slippage detection. The proposed sensor may be attached to a wide range of end-effectors, including both soft and rigid grippers. Some examples are listed as follows.[001 10] I) Agriculture and Food Processing: In agricultural and food processing industries, robotic grippers are used to contact with fragile products such as vegetables and eggs. The inclined soft ridges of this disclosed sensor keep soft contact with the object surface, which prevents scratching and damage of the object, while in the meantime, providing contact force and slippage alarm.[001 11] II) Medical Devices: In the medical field such as surgical robots, rehabilitation robots and prosthetics, accurate force measurement and slippage detection are critical for delicate procedures and patient safety. This disclosed sensor can be minimized and then utilized in surgical instruments in robotic-assisted surgery systems to provide precise force feedback and ensure optimal performance. In addition, the slippage sensor can be integrated into the fingersof prosthetic limbs to monitor slippage during object handling and provide early warnings. A potential application is delivering feedback through adjustable electrical stimulation, vibrations, or thermal effects. This feedback allows users to assess the likelihood of slippage based on the intensity of the current, vibration, or temperature change.

[0112] III) Manufacturing and assembly: In robotic manufacturing, precise control and feedback are crucial for efficient operation This disclosed sensor can equip robotic grippers with accurate force measurements and reliable slippage detection so that the gripper can monitor and adjust contact forces in real time, which ensures the safe and efficient handling of delicate or high-value components and reduces the risk of damage.

[0113] Referring to FIG. 33, in embodiments of the present disclosure, the processing system 900 may include a controller 901 and user interface 902. User interface 902 is configured to enable manual interactions between a user and the computing module as required. For this purpose, the processing system 900 includes the input / output components required for the user to enter instructions to provide updates to each of the modules. A person skilled in the art will recognize that components of user interface 902 may vary from embodiment to embodiment but may typically include one or more input devices 935 such as but not limited to a touchscreen, a keyboard, a joystick, a mouse, a microphone, etc. The user interface 902 can also include a media player 940, which can be in the form of one or more playback devices, including but not limited to a display, a speaker, earphones, headsets, etc.

[0114] The controller 901 is configured to be in data communication with the user interface 902 via bus 915. The controller 901 includes memory 920 and processor 905 mounted on a circuit board to process instructions and data, e.g., to perform the method of the present disclosure. The controller 901 includes an operating system 906, an input / output (I / O) interface 930 for communicating with user interface 902, and a communications interface, e.g., a network card 950. The network card 950 may, for example, be configured to send data from thecontroller 901 via a wired or wireless network to other processing devices or to receive data via the wired or wireless network. Wireless networks that may be utilized by the network card 950 include, but are not limited to, Wireless-Fidelity (Wi-Fi), Bluetooth, Near Field Communication (NFC), cellular networks, satellite networks, telecommunication networks, Wide Area Networks (WAN), and etc.[001 15] Memory 920 and operating system 906 are in data communication with central processing unit (CPU) 905 via bus 910. The memory 920 may include both volatile and nonvolatile memory. The memory 920 may include more than one of each type of memory, e.g., Random Access Memory (RAM) 923, Read Only Memory (ROM) 925, and a mass storage device 927. The mass storage device 927 may include one or more solid-state drives (SSDs). One skilled in the art will recognize that the memory described above includes non -transitory computer-readable media and shall be taken to include all computer-readable media except for a transitory, propagating signal. Typically, instructions are stored as program code in the memory but can also be hardwired. Memory 920 may include a kernel and / or programming modules such as a software application that may be stored in either volatile or non-volatile memory.

[0116] Herein, the term “processor” is used to refer generically to any device or component that can process computer-readable instructions, including for example, a microprocessor, microcontroller, programmable logic device, or other computational device. That is, processor 905 may be provided by any suitable logic circuitry for receiving inputs, processing them in accordance with instructions stored in memory, and generating outputs (for example to the memory components or media player 940). In the present disclosure, processor 905 may be a single core or multi-core processor with memory addressable space. In one example, processor 905 may be multi-core, comprising — for example — an 8 core CPU. In another example, it could be a cluster of CPU cores operating in parallel to accelerate computations.

[0117] Further, one skilled in the art will recognize that certain functional units in this description have been labelled as modules throughout the specification. The person skilled in the art will also recognize that a module may be implemented as circuits, logic chips or any sort of discrete component. Still further, one skilled in the art will also recognize that a module may be implemented in software which may then be executed by a variety of processor architectures. In embodiments of the disclosure, a module may also comprise computer instructions or executable code that may instruct a computer processor to carry out a sequence of events based on instructions received. In further embodiments, the module may comprise a combination of different types of modules or sub-modules. The choice of the implementation of the modules may be determined by a person skilled in the art and does not limit the scope of the claimed subject matter in any way.

[0118] All examples described herein, whether of apparatus, methods, materials, or products, are presented for the purpose of illustration and to aid understanding, and are not intended to be limiting or exhaustive. Modifications may be made by one of ordinary skill in the art without departing from the scope of the invention as claimed.

Claims

CLAIMS1. A sensor, comprising: a multilayer base defining a thickness axis, the multilayer base comprising: a substrate layer; a sensing layer coupled to the substrate layer, the substrate layer being more rigid than the sensing layer; at least one Fiber Bragg Grating (FBG) embedded in the sensing layer, wherein the substrate layer and the sensing layer are collectively bendable relative to a neutral plane, wherein the at least one FBG is spaced apart from the neutral plane along the thickness axis, and a plurality of projections coupled to a sensing surface of the sensing layer and adjacent to the at least one FBG, wherein the plurality of projections defines at least one contact region on the sensing surface.

2. The sensor as recited in claim 1, wherein responsive to the at least one contact region receiving a shear force along a shear sensing axis, the at least one FBG undergoes tension to shift a FBG wavelength of an optical signal along a first shift direction, wherein the shear sensing axis is transverse to the thickness axis.

3. The sensor as recited in claim 2, wherein responsive to the at least one contact region receiving a normal force along a normal sensing axis, the at least one FBG undergoes compression to shift the FBG wavelength of the optical signal along a second shift directionopposing the first shift direction, wherein the normal sensing axis is transverse to the shear sensing axis.

4. The sensor as recited in any of claims 2 and 3, wherein each of the plurality of projections is tilted relative to the thickness axis towards a tilt direction parallel to the shear sensing axis.

5. The sensor as recited in any one of claims 1 to 4, wherein each of the plurality of projections has a respective height along the thickness axis larger than a thickness of the multilayer base along the thickness axis.

6. The sensor as recited in any one of claims 1 to 4, wherein each of the plurality of projections has a respective height along the thickness axis smaller than a thickness of the multilayer base along the thickness axis.

7. The sensor as recited in any of the above claims, wherein each of a plurality of contact regions corresponds to a respective ones of a plurality of FBGs.

8. The sensor as recited in any of the above claims, wherein the plurality of projections comprises a plurality of first projections adjacent to a first FBG, and a plurality of second projections adjacent to a second FBG9. The sensor as recited in claim 8, wherein the plurality of first projections defines a first contact region, and the plurality of second projections defines a second contact region.

10. The sensor as recited in any one of claims 8 and 9, wherein the plurality of first projections are tilted relative to the thickness axis towards a first tilt direction; and the plurality of second projections are tilted relative to the thickness axis towards a second tilt direction, wherein the second tilt direction is opposite to the first tilt direction.1 1 . The sensor as recited in any one of claims 8 to 10, wherein the first FBG is parallel to the second FBG.

12. The sensor as recited in any one of claims 8 to 10, wherein the first FBG is transverse to the second FBG, and wherein the first FBG is offset from the second FBG along the thickness axis.

13. The sensor as recited in any of the above claims, wherein each of the plurality of projections is at least equally rigid or more rigid than the sensing layer.

14. The sensor as recited in any of the above claims, wherein the substrate layer comprises a pair of supporting legs coupled to opposing ends of the substrate layer.

15. The sensor as recited in any of the above claims, wherein the multilayer base further comprises: a second sensing layer coupled to the substrate layer, the sensing layer and the second sensing layer coupled to opposing surfaces of the substrate layer, the substrate layer being more rigid than the second sensing layer; at least one second FBG disposed embedded in the second sensing layer,wherein the substrate layer, the sensing layer and the second sensing layer are collectively bendable relative to the neutral plane, wherein the at least one second FBG is spaced apart from the neutral plane along the thickness axis and a plurality of second projections coupled to a second sensing surface of the second sensing layer, the plurality of second projections defining at least one second contact region on the second sensing surface.

16. A sensor system, comprising: at least one sensor as recited in any one of the above claims; and an optical interrogator in optical communication with each of the at least one sensor, the optical interrogator being configured to receive an optical signal from each of the at least one sensor.

17. The sensor system as recited in claim 16, further comprising a controller in signal communication with the optical interrogator, the controller configured to convert the optical signal from each of the at least one sensor into a respective sensing signal.

18. The sensor system as recited in claim 17, wherein the controller is configured to convert the optical signal from each of the at least one sensor into the respective sensing signal using a machine learning model.

19. The sensor system as recited in claim 18, wherein the machine learning model comprises at least one of: a Multilayer Perceptrons (MLPs), a Convolutional Neural Networks(CNNs), a Long Short-Term Memory networks (LSTMs), a Gated Recurrent Units (GRUs), and Transformer-based architectures.

20. A robotic gripper, comprising: at least one gripper finger; and at least one sensor as recited in any one of claims 1 to 15 coupled to the at least one gripper finger.

21. A prosthetic hand, comprising: at least one prosthetic finger; and at least one sensor as recited in any one of claims 1 to 15 coupled to the at least one prosthetic finger.

22. A method of sensing, comprising: a sensor as recited in any one of claims 1 to 15; receiving a shear force along a shear sensing axis on the at least one contact region of the sensor, such that the at least one FBG undergoes tension to shift a FBG wavelength of an optical signal along a first shift direction; and receiving a normal force along a normal sensing axis on the at least one contact region of the sensor, such that the at least one FBG undergoes compression to shift the FBG wavelength of the optical signal along a second shift direction, wherein the shear sensing axis is transverse to the normal sensing axis, and wherein the second shift direction is opposing the first shift direction.

23. The method as recited in claim 22, further comprising: converting an optical signal from each of the at least one sensor into a respective sensing signal.

24. The method as recited in claim 23, further comprising: converting the optical signal from each of the at least one sensor into the respective sensing signal using a machine learning model.

25. The method as recited in claim 24, wherein the machine learning model comprises at least one of: a Multilayer Perceptrons (MLPs), a Convolutional Neural Networks (CNNs), a Long Short-Term Memory networks (LSTMs), a Gated Recurrent Units (GRUs), andTransformer-based architectures.

Images