Tool arranged to manually operate a mechanical connector between an open state and a closed state

The tool with an image sensor and machine learning capabilities addresses the challenge of inaccurate closure detection in manual mechanical connectors by accurately identifying the connector's state, enhancing traceability in industrial assembly.

WO2026082336A1PCT designated stage Publication Date: 2026-04-23ATLAS COPCO IND TECHNIQUE AB INTELLECTUAL PROPERTY DEPARTMENT
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
ATLAS COPCO IND TECHNIQUE AB INTELLECTUAL PROPERTY DEPARTMENT
Filing Date
2025-09-01
Publication Date
2026-04-23

AI Technical Summary

Technical Problem

Existing manual tools for operating mechanical connectors, such as POPP clamps, lack reliable error proofing capabilities, leading to inaccurate traceability in industrial assembly processes due to challenges in detecting the correct closure of the connectors, especially when operators use varying techniques or the tools experience false detection from vibrations and strains.

Method used

A tool equipped with an image sensor, such as a Time-of-Flight sensor, captures images of the mechanical connector during operation to determine its state, using machine learning or image recognition techniques to accurately identify whether the connector is in an open or closed position, independent of the operator's technique or movement-induced vibrations.

Benefits of technology

Enhances error proofing by reliably detecting the correct closure of mechanical connectors, reducing false detections and improving the traceability of manufacturing processes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure EP2025074747_23042026_PF_FP_ABST
    Figure EP2025074747_23042026_PF_FP_ABST
Patent Text Reader

Abstract

A tool (1) is provided being arranged to manually operate a mechanical connector (6) between an open state and a closed state, wherein the mechanical connector (6) is arranged to join two pieces of material together in the closed state. The tool (1) comprises a tool tip (5) arranged to engage with the mechanical connector (6) so as to operate it; a sensor (10) arranged to produce an image of at least a portion of the mechanical connector (6) upon operation of the tool (1); and a control device (2) configured to estimate whether the mechanical connector (6) is in the closed state by recognizing whether the image from the sensor (10) represents the mechanical connector (6) in the closed state.
Need to check novelty before this filing date? Find Prior Art

Description

[0001]TOOL ARRANGED TO MANUALLY OPERATE A MECHANICAL CONNECTOR BETWEEN AN OPEN STATE AND A CLOSED STATEField of the inventionThe present invention generally relates to the field of tools for manuallyoperating mechanical connectors between an open state and a closed state for joining two pieces of material together. Background of the invention In industrial assembly, different kind of mechanical connectors are used tomanually join two or more pieces of material together. Examples of suchmechanical connectors are different types of clamps and clips, such as so called pre-opened pre-positioned clamps (in the following referred to as POPP clamps).For example, vehicles (or other products) may comprise hoses (flexible pipes)that may be applied to their fittings by means of mechanical connectors such as POPP clamps. These kinds of mechanical connectors are typically operated by entirely manual tools, i.e. tools that does not rely on any motorized power for operation. Instead, the operator makes the twisting / bending / pushing / pulling movement with the tool, which closes the mechanical connector. In industrial manufacturing, traceability of the manufacturing process steps isimportant to keep track of that all the process steps have been performedcorrectly. As tools for operating mechanical connectors of the above-mentionedtype are entirely manual and does not comprise any motor or torque sensor, it isdifficult to record whether the mechanical connector has been closed properly by operation of the tool.Tools for operating POPP clamps have been developed that are able to detectand record the closing (releasing) of a clamp. However, there is a risk that a false detection of a clamp release is made e.g. if an operator drops the tool on the floor or hits the tool against a hard surface. A production management system may then erroneously detect that a clamp has been released even if it has not,whereby the accuracy of the traceability of the manufacturing is reduced.Further efforts have therefore been made to develop such tools with improvederror proofing capabilities. Examples of such tools are disclosed inWO2024132305A1 and WO2024132304A1. Such tools rely on sensors such asstrain gauges and accelerometers as well as location techniques to estimatewhether the clamp has actually been correctly released. Many of these sensor types have their own challenges when it comes todetecting a signature event that indicates a closing of the connector. Hence, thereis still a need for improvements in the field. of the inventionIt would be advantageous to achieve a tool and a method overcoming, or at leastalleviating, the above-mentioned drawbacks. In particular, it would be desirableto enable a tool and a method with improved error proofing capabilities.To better address one or more of these concerns, a tool and a method having thefeatures defined in the independent claims are provided. Preferableembodiments are defined in the dependent claims.Hence, according to a first aspect, a tool arranged to manually operate amechanical connector between an open state and a closed state is provided. Themechanical connector is arranged to join two pieces of material together in theclosed state. The tool comprises: a tool tip arranged to engage with themechanical connector so as to operate it; a sensor arranged to produce an image of at least a portion of the mechanical connector upon operation of the tool; and a control device configured to estimate whether the mechanical connector is inthe closed state by recognizing whether the image from the sensor represents(such as depicts) the mechanical connector in the closed state.The inventor has realised that, when using other types of sensors (as in the priorart), it may be challenging to detect a signature event indicative of a trueconnector closing. For example, if a strain gauge is used to detect the twist / bending of the shaft of the tool upon operation, the amount of force / exactbending direction may be rather individual between different operators. Hence,it can be difficult to provide a signal analysis that can make correct detections of connector closings for any operator of the tool. Further, signals from accelerometers / gyros may comprise lots of peaks / patterns that mistakenly may be detected as a connector closing. Further, sometimes, even if the operator applies sufficient pressure to close the connector, it may not close correctly. However, the tool may incorrectly indicate that the connector is closed because it detected the appropriate strain or vibration. With the present aspect, error proofing is made possible that is not dependenton the actual closing action, i.e. the twisting / bending movement of the tool, orany vibration caused by the closing of the mechanical connector. Instead, an image of at least a portion of the mechanical connector is captured by the sensor,wherein the estimation of whether the connector is closed is based on the image.Thereby, the error proofing of the present aspect is less influenced by individual operation techniques among the operators and also by different mechanical connector behaviours during closing. Accordingly, a tool with improved error proofing capabilities is provided.Images may be captured by the sensor and analysed by the control devicecontinuously (i.e. at a predetermined frequency) during usage of the tool (suchas at east during movement of the tool or at least when the tool is powered on)or at least just as the tool has operated the mechanical connector.That the tool is arranged to manually operate the mechanical connector means that the operation of closing the mechanical connector by the tool is completelymanual, without any aid of a motor / machine. Hence, the tool is un-motorized.The sensor may e.g. be arranged at the tool (e.g. on a handle of the tool) so as to produce an image of an area just beside / around the tool tip, preferably such that a portion of the mechanical connector is captured in the image (upon operationof the tool), which portion’s position and / or configuration characterizes theclosed state of the connector. In the present specification the term “image” is to be broadly interpreted as any at least two-dimensional digital representation of the imaged object as captured by the sensor. Hence, the image may not necessarily be a visual representation that the human eye would perceive as a true rendering of the imaged object. According to an embodiment, the image may be a 3D image (which also may be referred to as a depth image), wherein each one of several zones of the image comprises information related to the distance from the sensor to the imaged object. For example, the sensor may be a depth image sensor.For example, one zone may be one pixel or one group of pixels of the image.For example, the image may be a matrix, the elements of which represents the sensed distances across the field of view of the sensor.The present embodiment is advantageous in that it facilitates processing of theimage for the control device in order to estimate if the connector is closed. The distance information across the field of view of the sensor is relatively small in(data) size, yet still informative of the position / configuration of the connector.A 2D image (without any information related to the distance to the imagedobject) may alternatively be envisaged but may require more processing powerfor the control device to recognize a closed connector. For example, the sensor may be a Time-of-Flight (ToF) sensor. ToF sensors are generally relatively inexpensive which thus may lower the product cost. Other depth / 3D sensors may alternatively be envisaged.The recognition of if the image represents the mechanical connector in theclosed state may be based on reference image data (e.g. comprising a plurality ofimages) representing (such as depicting) mechanical connectors in a closedstate. The reference image data may e.g. be used for training a machine learning model or may be used for (e.g. direct) comparison with the image captured by the sensor in case using other image recognition techniques. According to an embodiment, the estimation may be made based on a machinelearning model trained on images representing closed mechanical connectorsand images not representing closed mechanical connectors. For example, themodel may be trained on images captured as the tool is engaged with a closed connector and images captured as the tool is not engaged with a closed connector, for example images when the tool is just held in the air or images taken when the tool is engaged with an open mechanical connector. Alternatively (or as a complement), the estimation may be made based on any other suitable image / pattern recognition technique, such as normalizes Cross- Correlation (NCC), Structural Similarity Index (SSIM) or Cross Correlation. The mechanical connector may e.g. be of any kind that has a visually distinctive discrepancy between its’ opened and closed states, thereby allowing the sensor and control device to distinguish between the two states. According to an embodiment, the mechanical connector may be a pre-opened pre-positioned (POPP) clamp. Prior to mounting, POPP clamps are pre-tensioned in an open position, in which the diameter of the clamp is slightly wider than in the clamp’s closed position. A protrusion in one end of the clamp band may abut a portion of the other end of the clamp band so as to hold the clamp under tension in the open position. When the protrusion is forced out of abutment with the other end, the clamp is released and snaps to its closed position with a narrower diameter, whereby the clamp will secure the hose to its fitting. POPP clamps are advantageous to use in industrial manufacturing, as their installation during the assembly process is fast. The tool may be used to force the protrusion out of abutment with the other end of the clamp band and may will only take a single twist / bending movement of the tool to release the clamp. Examples further to POPP clamps are clips, hose clamps etc. According to an embodiment, the control device may be further configured toissue an indication in case it is estimated that the mechanical connector is in theclosed state. For example, such an indication may be sent to a central control system and / or to an indicator device (such as a display and / or LED) of the tool, that may signal to the operator that the connector is properly closed. According to an embodiment, the control device may be further configured to estimate whether the mechanical connector is in the open state by recognizingwhether the image from the sensor represents (such as depicts) the mechanicalconnector in the open state. Hence, the tool may also be able to detect if the mechanical connector is in the open state.According to a second aspect a method performed by a control device of a tool isprovided. The tool is arranged to manually operate a mechanical connector between an open state and a closed state, wherein the mechanical connector is arranged to join two pieces of material together in the closed state, the tool comprising a tool tip arranged to engage with the mechanical connector so as to operate it and a sensor arranged to produce an image of at least a portion of the mechanical connector upon operation of the tool. The method comprises estimating whether the mechanical connector is in the closed state byrecognizing whether the image from the sensor represents the mechanicalconnector in the closed state. It is noted that embodiments of the invention relate to all possible combinations of features recited in the claims. Further, it will be appreciated that the variousembodiments described for the tool are all combinable with the method asdefined in accordance with the second aspect of the present invention. Brief description of the drawingsThis and other aspects will now be described in more detail in the followingillustrative and non-limiting detailed description of embodiments, withreference to the appended drawings. Figure 1 shows a tool according to an embodiment. Figure 2a shows the tool and a mechanical connector in an open state. Figure 2b shows the tool and a mechanical connector in a closed state.Figure 3a-3c shows images captured by a sensor of the tool according to anembodiment. All the figures are schematic, not necessarily to scale, and generally only show parts which are necessary in order to elucidate the embodiments, wherein otherparts may be omitted. Like reference numerals refer to like elements throughoutthe description. Detailed description of embodimentsA tool 1 according to an embodiment will be described with reference to Figure1. The tool 1 is arranged to operate mechanical connectors 6, such as POPP clamps (as illustrated in Figure 1), during an industrial assembly process. The POPP clamp may have a circular band that is pre-tensioned in an open state. The tool 1 can be operated to pull a loop 7 at the end of the band, whereby the POPP clamp releases and snaps into its closed state. Typical objects to which POPP clamps can be applied are hoses to be secured to their fittings.The tool 1 may comprise a tool body 8 including a handle 3 and a shaft 4. Thetool 1 comprises a tool tip 5 (which also may be referred to as a tool head), e.g.arranged at the end of the shaft 4. The tool tip 5 is arranged to engage with themechanical connector 6. Optionally, the tool tip 5 may be exchangeable to othertool tips adapted to operate other kinds of mechanical connectors. The tool 1 comprises a sensor 10 arranged to produce an image of at least aportion of the mechanical connector 6. For example, the sensor 10 may bearranged such that, when the tool tip 5 engages the mechanical connector 6 foroperation thereof, it’s field of view is aimed at a portion / feature of themechanical connector that will change its position / character as the connector 6 closes. An example of such a portion is the feature 14 of the POPP clamp, thatwill change its position as the clamp closes. For example, the sensor 10 may bearranged so as to aim towards the area around / just beside the tool tip 5. For example, the sensor 10 may be arranged at the handle 3 of the tool 1. The sensor 10 may preferably be a 3D sensor, wherein the image comprises distance information across the field of view of the sensor 10. For example, the sensor 10 may be a ToF sensor. Other types of image sensors may also be envisaged. For example, the image produced by the sensor 10 may comprise a matrix, such as a 8x8 matrix, wherein each element of the matrix comprises distance information from the sensor 10 to the associated portion of the imaged object. The tool 1 comprises a control device 2. The control device 2 may be arrangedinside the tool body 8 (such as in the handle 3 as illustrated in Figure 1) orremote from the tool body 8 (such as in a central computer) and connected tothe tool body 8 wirelessly or via a cable. The control device 2 may comprise amemory and processing means. The control device 2 may be configured to receive and process image data fromthe sensor 10. The control device 2 is configured to estimate whether themechanical connector 6 is in the closed state based on the image from the sensor10. In particular, this is made by recognizing whether the image from the sensor10 is representing the mechanical connector 6 in the closed state. For example,the control device 2 may be configured to perform image recognition on theimage produced by the sensor 10 in order to estimate if the connector 6 is in theclosed state. This may e.g. be made by means of a machine learning model trained on images representing a mechanical connector 6 in at least in the closedstate. The machine learning model may also be trained on images representingthe mechanical connector being in an open state and / or images taken when the tool is held in the air. Other image / pattern recognition techniques may also be envisaged.Optionally, the control device 2 may be configured to also estimate whether themechanical connector 6 is in the open state by recognizing whether the imagefrom the sensor 10 is representing the mechanical connector 6 in the open state. For example, the control device 2 may be configured to start monitoring theimages produced by the sensor 10 in response to a trigger event, such as if anadditional sensor (such as a strain gauge and / or accelerometer / gyro) of the tool1 detects that the tool 1 makes a movement indicating that it operates amechanical connector 6. This may save computing power as well as reduce the risk of false detection of a closed connector 6. Alternatively, the control device 2 may be configured to start monitoring theimages produced by the sensor 10 as soon as the tool 1 is powered on or as arandom movement of the tool 1 is registered (e.g. by an accelerometer).Optionally, the control device 2 may use data from other sensors (such as straingauges and accelerometers) and / or location technique (as disclosed in any one of WO2024132305A1 and WO2024132304A1) to further confirm if the mechanical connector 6 is in the closed state.In the following, it will be described how the tool 1 functions with reference toFigures 2a and 2b. First, the operator engages the tool tip 5 with the openmechanical connector 6 (Figure 2a). When the tool 1 is held in this position, thesensor 10 registers the feature 14 of the connector 6 on a certain distance. Thenthe operator bends the tool 1 so the tool tip 5 pushes the feature 14, whereby themechanical connector 6 snaps into its closed state (Figure 2b). When the tool 1 isin this position, the sensor 10 registers the feature 14 on a closer distance. Thecontrol device 2 processes the image taken in this position of the tool 1 andrecognises the image as representing the mechanical connector 6 in the closedstate. The control device 2 may then accordingly estimate the mechanicalconnector 6 to be in the closed state. In Figures 3a to 3c, three example images are illustrated, that are produced bythe sensor 10 in three different situations. Each image comprises an 8x8 matrix,wherein each element of the matrix comprises distance information (in mm) tothe imaged object (with a maximum registerable distance limited 100 mm).Figure 3a is an image produced when the tool 1 is simply held in the air and isnot engaged with / close to the mechanical connector 6. As can be seen, several ofthe matrix elements show the maximum distance. Figure 3b is an imageproduced when the tool 1 is engaged with an open mechanical connector 6,while Figure 3c is an image produced when the tool 1 is engaged with a closedmechanical connector 6. As can be seen, several of the central elements of thematrix of Figure 3c (closed mechanical connector 6) shows shorter distanceinformation as compared to the central elements of the matrix of Figure 3b (open mechanical connector 6). The control device 2 may thus distinguish between the different images and estimate the state of the mechanical connector 6 accordingly. The person skilled in the art realizes that the present invention by no means is limited to the embodiments described above. On the contrary, many modifications and variations are possible within the scope of the appended claims. Additionally, variations to the disclosed embodiments can be understood and effected by the skilled person in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that acombination of these measures cannot be used to advantage.

Claims

CLAIMS1. Tool (1) arranged to manually operate a mechanical connector (6)between an open state and a closed state, wherein the mechanical connector (6) is arranged to join two pieces of material together in theclosed state, the tool (1) comprising:a tool tip (5) arranged to engage with the mechanical connector(6) so as to operate it;a sensor (10) arranged to produce an image of at least a portion ofthe mechanical connector (6) upon operation of the tool (1); anda control device (2) configured to estimate whether themechanical connector (6) is in the closed state by recognizing whetherthe image from the sensor (10) represents the mechanical connector (6)in the closed state.

2. Tool (1) according to claim 1, wherein the image is a 3D image, whereineach one of several zones of the image comprises information related to the distance to the imaged object.

3. Tool (1) according to claim 2, wherein the sensor (10) is a Time-of-Flightsensor.

4. Tool (1) according to any one of the preceding claims, wherein saidestimation is made based on a machine learning model trained on imagesrepresenting closed mechanical connectors (6) and images notrepresenting closed mechanical connectors (6).

5. Tool (1) according to any one of the preceding claims, wherein themechanical connector (6) is a pre-opened pre-positioned (POPP) clamp.

6. Tool (1) according to any one of the preceding claims, wherein thecontrol device (2) is further configured to issue an indication in case it isestimated that the mechanical connector (6) is in the closed state.

7. Tool (1) according to any one of the preceding claims, wherein thecontrol device (2) is further configured to estimate whether themechanical connector (6) is in the open state by recognizing whether theimage from the sensor (10) represents the mechanical connector in theopen state.

8. Method performed by a control device (2) of a tool (1) arranged tomanually operate a mechanical connector (6) between an open state anda closed state, wherein the mechanical connector (6) is arranged to jointwo pieces of material together in the closed state, the tool (1) comprisinga tool tip (5) arranged to engage with the mechanical connector (6) so asto operate it and a sensor (10) arranged to produce an image of at least a portion of the mechanical connector (6) upon operation of the tool (1),the method comprising: -estimating whether the mechanical connector (6) is in the closedstate by recognizing whether the image from the sensor (10) representsthe mechanical connector (6) in the closed state.

Citation Information

Patent Citations

  • Hose clamp tool

    US20050262969A1

  • Control device and method for a tool arranged to release pre-opened pre-positioned (POPP) clamps

    WO2024132304A1

  • Tool arranged to release pre-opened pre-positioned (POPP) clamps

    WO2024132305A1