GB2486658A

GB2486658A - Torque tool positioning system

Info

Publication number: GB2486658A
Application number: GB1021627.3A
Authority: GB
Inventors: Neil Mcdonald
Original assignee: Crane Electronics Ltd; Crane Electronics Inc
Current assignee: Crane Electronics Ltd; Crane Electronics Inc
Priority date: 2010-12-21
Filing date: 2010-12-21
Publication date: 2012-06-27
Also published as: GB201021627D0

Abstract

The system is for detecting the position of a manually positionable torque tool 14 in relation to e.g. fasteners 8a-8d of a work piece 2. The tool 14 includes a head 15 for engagement with a rotary fastening and a light source 18 mounted on the head for indicating the position of the head to a remote position detection system. The intensity of light emitted by the light source may be varied in a pattern such as a series of pulses. The light source may be infra-red. A system and method for detecting the position of a torque tool includes a camera 10 arranged to view a work area 4 and being sensitive to light emitted by a light source on the tool. A memory stores a set of locations within the field of view of the camera, the locations corresponding to a set of operating positions in the work area. Also included is a means 12 for generating a coincidence signal when the sensed location of the light source on the tool coincides with one of the stored locations.

Description

TITLE

Torque tool positioning system

DESCRIPTION

Technical field

The invention relates to manufacturing and assembly processes that entail the use of manually positioned torque tools to apply measured levels of torque to rotary fastenings. It has particular application to processes in which multiple fastenings must be tightened in a particular sequence or to different torque levels.

Background of the invention

In industries that use nuts and bolts or other rotary fastenings, not only is it important that the correct torque is applied, but it is also important that this torque is applied to the correct fastening. This concept becomes more important where there are several fastenings that make up the overall joint between two surfaces. For example, a wheel can require 4 or 5 nuts to be tightened up correctly for it to be safely attached to a wheel axle. Another example is an engine where the cylinder head is fastened down with 8 bolts, which have to be tightened in a certain sequence to avoid warping the surface. Another example is a circular flange, which might be fastened with 60 bolts, in a clockwise sequence, with the additional difficulty that the torque is increased in 3 stages, so that all bolts are tightened up to a first torque, the process is repeated to take them all up to a second torque and then the process is repeated again taking the fastenings up to the required final torque.

Even if the sequence in which the fastenings are tightened is unimportant, the required torque niay vary between the different fastenings so it is important to apply the correct torque to the respective fastenings.

Modem torque tools are often in electronic communication with a computer, which records torque readings taken by the tool in order to create a log of the assembly of each component, which can be useful for ensuring quality and safety. Clearly it is important that the correct reading is logged against the correct fastener.

One possibility for ensuring the correct sequence of operation is to automate the task using a robot arm, which could be programmed to go to through a set of fastenings in a certain order. However, there are problems with a robotic solution. Robots are expensive and a lot of time is required to ensure they are programmed correctly not just to tighten the correct fastening but to make sure that the fastening tool enters and exits the work area in a safe manner, and can be moved in an appropriate fashion to actually tighten the joint. Sufficient clearance must be provided for the arm mechanism. If the joint is a hose or pipe for attachment to an engine assembly, for example, then the fastening tool will need to have an open wrench end and so cannot be just part of a rotary tool. Also, owing to our lack of trust in robots, they normally have to be encased in a safety cage to prevent human access while they are operating, which adds further to the cost.

Alternatively, a human operator could be trained to follow a certain process of tightening. They would then rely on their memory to perform the tightening sequence correctly. This process could be aided by pictures in folders or on screens in the work area. The operator could also dab a coloured ink onto each tightened bolt to confirm it was already tightened. However relying on a person no matter how experienced they are can also be prone to problems. They can forget where they are in the sequence, especially if working on the same type of joint over and over again in an assembly line. To record or mark joints as they are tightened would help to improve matters but would slow down the assembly process.

A workpiece comprising a joint that requires tightening is typically presented in a standard position and orientation in the work area. Therefore, if it were possible to sense automatically the position of a manually operated tool, it would be possible to determine which fastening the tool was engaged with and feedback could be given to the operator to ensure that the correct sequence of operations was followed. The difficulty of this is compounded because the tool position has six degrees of freedom.

The tool has a shape with a head at one end that typically receives one of various torque fastening adaptors, and a handle at the other end for the operator to hold the tool and pull torque. Three of the degrees of freedom relate to the linear position of the tool in three dimensions (x,y,z) within the work area. There are three further degrees of freedom (which in one co-ordinate system can be characterized as yaw, pitch and roll) relating to the orientation of the tool while a reference point such as its centre of gravity remains at a fixed linear position.

Various systems have been proposed for working out the position of the tool within the work area. Some examples have proposed using a form of positioning system based on the travel time of signals -for example, ultrasonics -from fixed beacons around the work area. These can suffer from two problems. First, they cannot be precise when fastenings are positioned close together. Second, they cannot take into account the orientation of the tool. Thus there can arise a situation in which the system senses the location of the receiver inside the tool and determines that the tool is correctly positioned but in fact, because of a wrong orientation, the handle is next to the desired fastening or, worse, the head of the tool is actually positioned over an adjacent fastening. These approaches also require the tool, which needs to be kept small, to have extra receiver circuitry (uhrasonic or GPS) built into it. They also rely on a good coverage in the work area without signal reflections that may cause distance calculation errors.

Another approach has been to use visual recognition, where a computer would study a picture of the work area received from one or more cameras, would identify the fastenings and the tool as it was being positioned and would alert the operator when the correct part of the tool was positioned over the correct fastening. The problem here is that the fastening and the parts of the joint to be fastened can be made from similar materials and so present a low contrast to the visual system, making them hard for the computer to identify reliably. In factories it would be hard to achieve a good contrast and lighting level in all parts of the work area. The hand of the operator will necessarily obscure part of the tool, which may make its shape hard for the system to recognize, and the operator's body may hide part of the work area from the view of some of the cameras, or cast confrising shadows across it. The information from two cameras would typically be converted into a three-dimensional, stereoscopic representation of the work area, which cannot be done if the view of one of the cameras is obstructed. The algorithms used in visual recognition systems are generally complicated and prone to error.

Summary of the invention

The invention provides a torque tool as defined in claim 1; a system for detecting the position of a torque tool as defined in claim 9; and a method of detecting the position of a torque tool as defined in claim 21.

Preferred but non-essential features of the invention are defined in the dependent claims.

The invention ensures that a torque tool is correctly manually positioned by an operator before use. Unlike the receivers of previously proposed positioning systems, a light source such as an LED is small enough to be mounted directly on the head of the torque tool so that the position of the light source necessarily coincides with the position of the fastening to which the tool is applied. This simplifies the image processing because the orientation of the handle of the tool becomes irrelevant.

Moreover the colour of the light source is well defined and can be chosen to contrast well with its surroundings, while the camera can be made sensitive to that particular colour. Therefore the light source can be easily distinguished from its background, further simplifying the image processing that is required and avoiding the need for special lighting within the work area. The system does not need to reconstruct a three-dimensional view of the work area, even if multiple cameras are used, but merely to compare the sensed location of the light source against a set of locations that are stored in its memory.

The invention is applicable to any manually positionable torque tools such as torque wrenches and nutrunners. Typically those tools will be held in the operator's hand.

However, the term "manually positionable" is intended to include tools that are supported by machinery but guided into position by the operator, or in which the guidance is automatic but it is the operator that controls the sequence in which fastenings are tightened by the tool.

The invention can be applied to simple torque tools with no display or in-built processor but its utility is extended when applied to "intelligent" digital wrenches.

The drawings Fig. 1 is a schematic overview of a system for detecting the position of a torque wrench according to the invention.

Fig. 2 represents the different views of the two cameras shown in Fig. 1.

Fig. 3 represents the view of a camera in a variant of the system in which the workpiece moves along a predetermined course.

Figure 1 is a schematic view, which is not to scale, representing a system for detecting the position of a torque wrench in accordance with the invention. A workpiece 2 is placed at a known position and in a known orientation within a work area 4. The workpiece 2 comprises two components 5,6 that need to be joined together using a set of fastenings, namely the four bolts 8a,8b,8c,8d. In a real example, the four bolts 8a-8d might need to be fastened in a specific sequence and might have differing requirements as to the torque with which they are fastened. The bolts are illustrated for simplicity as lying in a straight line but that typically will not be the case and indeed they need not even lie in the same plane.

The work area 4 lies within the field of view of two digital cameras 10,11 which are spaced apart from one another so that they have different views of the work area 2.

Figure 2 represents those respective views 20,21. In this example, the first caniera 10 is generally face on to the workpiece 2 and its view 20 is therefore generally similar to that shown in Fig. 1, i.e. the positions of the fastenings 8a-8d in the work area 4 are represented by a diagonal row of locations 22a-22d in the field of view 20. The second camera 11 is set at an angle to the workpiece 2 and in this example it happens to be generally aligned with the row of fastenings 8a-8d so that in the view 21 of the second camera 11 the positions of the fastenings 8a-8d in the work area 4 are represented by a foreshortened vertical row of locations 23a-23d.

The use of two cameras 10,11 allows a better depth of field to be achieved and reduces the risk of error that might occur if two fastenings appeared close to each other from the viewpoint of a single camera. In some cases only a single camera may be required, if it can be guaranteed that all of the fastenings 8a-8d will remain distinct and visible to that camera when they are tightened by an operator. In other cases more than two cameras may be required to deliver additional views, for example if the shape of the workpiece 2 means that some of the fastenings 8a-8d are hidden from certain viewpoints. The outputs of all of the cameras are fed into a computer 12 for processing.

A torque wrench 14 comprises, as is conventional, a head 15 at one end and a handle 16 at the opposite end. The wrench 14 may further comprise a screen 17 for displaying torque readings and other information to the operator. The head 15 may be shaped such that it can directly engage with fasteners 8a-8d but more typically the head 15 is equipped with a mounting that can receive any of a range of adapters (not illustrated) in order that the wrench 14 can apply torque in a controlled manner to a variety of different types of fastening.

The torque wrench 14 carries a bright LED source 18 located on its head 15, preferably directly over its adaptor so that the position of the LED 18 is for practical purposes the same as the position of the adaptor and the same as the position of any fastening with which the adaptor is engaged. The LED 18 has a specific colour, for example, red. The colour may be one that is invisible to a human eye but visible to the cameras 10,11, for example, infra red. The colour of the LED 18 should preferably be chosen to give good contrast with the background and in particular with the workpiece 2 so that it can be easily seen by the cameras 10,11. If the workpiece 2 has a colour that is similar to the original colour of the LED 18, then a different colour for the LED 18 and camera filters can be chosen. Alternative colours of LED 18 could be built into the wrench 14 with means provided in the wrench 14 for switching between them. Light sources other than LEDs are possible, though they are likely to be less efficient. If a white light source is used, it is preferably provided with a colour filter so that the light source can easily be recognized by the position detection system.

Filters may be used to enhance the contrast by reducing the sensitivity of the system to light frequencies other than the main ones emitted by the LED 18. Such filters can be physical filters (not illustrated) at the objective lenses of the cameras 10,11 or can be electronic filters in the processing carried out within the cameras 10,11 or the computer 12. The consequence is that each of the cameras 10,11 will see a bright spot corresponding to the position of the LED 18 and little else in its field of view. No complex visual processing is required and no attempt is made to recognize the shapes of objects or their three-dimensional relationships and orientations. Therefore the lighting and contrast in the work area 2 does not matter to the system and it can be optimized for the user or for efficiency as desired. The cameras 10,11 do not need to be complicated or expensive as they do not need to work in bad light conditions or with good contrast. They only have to detect one bright light source 18.

The computer 12 is preferably in data communication with the torque wrench 14. The wrench 14 could be connected to the computer 12 by a cable (not illustrated) but more preferably the system uses a wireless link via a transmitter/receiver 19 so that the wrench 14 can easily be positioned anywhere within the work area 4. Any suitable system of wireless communication can be used, for example Bluetooth®, WiFi® or RF. The LED 18 may receive its power from an external power source if the wrench 14 is attached by a cable or, more preferably, from an intemal power source, for

example, a battery.

Unlike a robot arm or a full visual recognition system, the training of the system according to the present invention is very simple. The workpiece 2 is placed in the work area 4. The head 15 of the wrench 14 is positioned over each fastening 8a-8d in turn and the LED source 18 is illuminated so that the position of that fastening can be viewed by each camera 10,11. The locations 22a-22d, 23a-23d that correspond to that fastening 8a-8d in the respective cameras' fields of view 20,2 1 are then recorded.

Either the LED 18 may be illuminated by a signal from the operator only when the wrench 14 is in position, which triggers the computer to record the locations; or the LED 18 may remain continuously illuminated and the operator may give a signal to the computer to record the locations after having positioned the wrench 14.

Each location will be stored in the memory of the computer 12 (or in any convenient location electronically accessible to the computer 12) as a pair of x and y co-ordinates for each camera, associated with a number identifying the fastener in question. If the fasteners need to be tightened in a particular sequence, the system will typically be trained by moving through the set of fasteners 8a-8d in that sequence. However, it is possible to store the locations of the fasteners in any order and to program the computer with the desired sequence afterwards. Other information may be stored in the computer and associated with each fastener, such as the torque required or the type of torque wrench that should be used with that fastener.

When an operator comes to use the wrench 14, they place the wrench adaptor over the fastening 8a-8d that they wish to tighten. They operate the LED source 18 or (if the LED 18 is continuously illuminated) they send a signal to the computer 12, and the computer then receives views of the work area 4 from the respective cameras 10,11.

The sensed location of the image of the LED 18 in each view 20,21 is compared with the stored locations of the fasteners 8a-8d in that view. If a match is found (i.e. the respective locations coincide within pre-deirned levels of tolerance) a determination may be made that the position of the head 15 of the wrench 14 coincides with the position of one of the fasteners 8a-8d and a coincidence signal may be generated by the computer 12. If multiple cameras are used, then a choice of algorithms for generating the coincidence signal is possible. The signal may be generated if a match is found in the field of view of any of the cameras 10,11. This would allow the system to operate successfully even if the fastener 8 in question is hidden from the viewpoint of one of the cameras 10,11. Altematively, the signal may be generated only if a match is found in the fields of view of all the cameras 10,11 (and if they agree on the identity of the fastener 8 in question). This would reduce the risk of errors in identification. With more than two cameras, further options between these two extremes could be considered, such as accepting a determination that the majority of cameras agree on.

The generation of a coincidence signal by the computer 12 signifies that generates a signal indicating that the fastener can be tightened. Preferably the computer 12 sends a signal to activate the wrench 14, which is disabled from taking torque measurements until that signal arrives. Alternatively, the signal sent by the computer 12 to the wrench 14 may simply cause the wrench 14 to provide an indication to the operator that the fastener may be tightened. That indication could be an audible indication such as a beep; a visible indication such as a message on the screen 17, the illumination of a dedicated LED (not illustrated) or the flashing of the head-mounted LED 18 in visible light (as the operator's gaze is already focussed in this direction); or a tactile indication created by vibration of the wrench 14 in the hand of the operator.

Even if the computer 12 is not in communication with the wrench 14, an audible or visible indication could be given to the operator by the computer itself 12.

The system just described is especially useful when the fasteners need to be tightened in a particular sequence. After training, the computer 12 has stored in its memory the location co-ordinates of each fastener 8, together with the number in the sequence of that fastener. When the system has recognized that position of the wrench 14 coincides with the position of a fastener 8, it may further make a determination whether that fastener 8 is the next in the sequence to be tightened and the computer 12 may only generate the coincidence signal if the outcome of that determination is true.

In the event that the fastener over which the wrench 14 is positioned is not the next one in sequence, the computer may simply fail to activate the wrench 14 or give the aforementioned positive indication to the operator; but more usefully it may generate a warning signal, alarm or message to identify the error.

The computer 12 may store other parameters associated with each fastener 8, in addition to or instead of its numerical sequence, which must also be verified before the coincidence signal can generated. Such parameters might include the type of -10 -torque wrench 14 or torque wrench adaptor that needs to be used with the fastener. If a communication link 19 is in place between the computer 12 and the wrench 14, that information could be transmitted from the wrench 14 as electronic data in a conventional manner. Alternatively, the necessary information could be transmitted by causing the LED 18 on the wrench 14 to turn on and off in predetermined patterns that can be detected by the cameras 10,11 and interpreted by the computer 12. Most simply, the LED 18 could be programmed to pulse at a frequency that indicates a characteristic (e.g. the size) of the wrench 14, with different types or sizes of wrench using different frequencies. If one wrench can carry different adaptors, the pattern of light emitted by the LED 18 could be changed to inform the computer 12 which adaptor was in use and the computer 12 could check its memory to verify that that adaptor was appropriate to the fastener 8 where the wrench 14 was positioned.

Instead of a continuous stream of pulses at a single frequency, the LED 18 might alternatively be turned on and off in more complex patterns to encode more complex information.

If the computer 12 is in two-way communication with the wrench 14, the computer 12 will typically record a log of the torque readings that are measured by the torque wrench 14 for the respective fasteners 8a-8d. In order that the correct readings could be logged against the correct fasteners, in the past this might require the operator to tighten the fasteners in a particular sequence, even if there was no engineering requirement for that sequence. Because the system of the present invention can identify each fastener 8 by its position, the readings measured by the wrench 14 can be reliably allocated to the respective fasteners 8, irrespective of the sequence in which the operator chooses to tighten them. Instead of the wrench 14 transmitting its torque readings to the computer 12 for storage there, the computer 12 could transmit to the torque wrench 14 the identity of the fastener 8 being tightened, so that the log can be created in memory on board the wrench 14 for later download.

The system so far described relies on the workpiece 2 being placed at a known position and in a known orientation within the work area 4. However, the same principle can be applied if the workpiece 2 (in a known orientation) moves on a predetermined course through the work area, for example as part of an assembly line.

In this case the training would take slightly longer. Each fastening 8a-8d would have the wrench 14 attached to it and the LED 18 would be illuminated continuously while the workpiece 2 moved along its course through the work area 4. The fastener 8a-8d would follow a fixed path, which would be detected by each camera 10,11 as a trace 25a-25d across its field of view 20,21, as shown in Fig. 3. The trace would be typically be a straight line but might be curved. Instead of the stored location of each fastener being a fixed point, the trace 25 would be stored in computer memory. It might be stored as a set of points but more efficiently the memory could store an equation representing or approximating the shape of the trace in the (x,y) co-ordinates of each camera's field of view. The computer 12, when it received the location of the LED 18 within the field of view of each camera 10,11 would calculate whether that location fell upon any of the traces 25a-25d stored within its computer memory.

If the workpiece 2 does not come into the work area 4 in a repeatable position and orientation, one or more reference LEDs (not illustrated) could be attached to known points on each workpiece 2. The system could then reference the positions of all fixings 8 on the workpiece relative to the detected locations of these reference sources. The reference LEDs would need to be detectable by the cameras 10,11 but distinguishable from the LED 18 on the wrench, e.g. by being a different colour or by flashing at a different frequency. The reference LED may be caused to switch on and off in a pattern that is detectable by the cameras 10,11 and can be interpreted by the computer 12, for example to provide the computer 12 with information about the type of workpiece 2 or its build standard.

Claims

-12 -CLAIMS1. A manually positionable torque tool comprising: a head for engagement with a rotary fastening; and a light source mounted on the head for indicating the position of the head to a remote position detection system.
2. A torque tool according to claim 1, wherein the intensity of light emitted by the light source is varied in a pattern that indicates a characteristic of the torque tool.
3. A torque tool according to claim 2, wherein the pattern is a series of pulses at a frequency that indicates the characteristic of the torque tool.
4. A torque tool according to claim 2 or claim 3, wherein the head can receive different adaptors for different rotary fastenings, and wherein the pattern indicates the type of adaptor currently received in the head.
5. A torque tool according to any preceding claim, further comprising means for receiving a signal from the position detection system to indicate when the head of the torque tool is in a correct operating position.
6. A torque tool according to claim 5, including means for disabling operation of the torque tool when the head is not in the correct operating position.
7. A torque tool according to claim 5 or claim 6, including means for indicating to an operator of the torque tool when the tool is in the correct operating position.
8. A torque tool according to any preceding claim, wherein the light source is an infra-red light source.
9. A torque tool according to any preceding claim, which is a hand-held torque tool.

-13 -
10. A system for detecting the position of a manually positionable torque tool, comprising: a camera arranged to view a work area, the camera being sensitive to light emitted by a light source on the torque tool; a memory for storing a set of locations within the field of view of the camera, the set of locations corresponding to a set of operating positions in the work area; and means for generating a coincidence signal when the sensed location of the light source on the torque tool coincides with one of the stored locations.
11. A system according to claim 10, further comprising a filter to reduce the sensitivity of the system to light of frequencies other than those emitted by the light source on the torque tool.
12. A system according to claim 10 or claim 11, wherein each of the locations stored in the memory comprises a trace across the field of view of the camera, the trace corresponding to a path followed by an operating position on a workpiece as the workpiece is moved along a predetermined course through the work area.
13. A system according to any of claims 10 to 12, wherein the memory further stores a defined sequence of the locations, and wherein the coincidence signal is generated only when the sensed location of the light source on the torque tool coincides with the next in sequence of the stored locations.
14. A system according to any of claims 10 to 13, further comprising means for detecting patterns in the intensity of light received by the camera from the light source.
15. A system according to claim 14, wherein each of the locations stored in the memory has associated with it a stored torque tool characteristic, each of the stored torque tool characteristics being represented by a pattern of light intensity, -14 -and wherein the coincidence signal is generated only when the detected pattern of light intensity represents the torque tool characteristic that is associated with the sensed location.
16. A system according to any of claims 10 to 15, further comprising means for transmitting the coincidence signal to the torque tool.
17. A system according to any of claims 10 to 16, wherein the camera is also sensitive to light emitted by a reference light source on a workpiece in the work area; and wherein the means for generating a coincidence signal makes use of locations relative to the sensed location of the reference light source.
18. A system according to any of claims 10 to 17, comprising a plurality of cameras arranged to have different views of the same work area, wherein for each camera the memory stores a set of locations within the field of view of that camera, the set of locations corresponding to the set of operating positions in the work area.
19. A system according to claim 18, wherein the coincidence signal is generated when the location of the light source sensed by any one of the cameras coincides with one of the stored locations for that camera.
20. A system according to claim 18, wherein the coincidence signal is generated only when the location of the light source sensed by each of the cameras coincides with one of the stored locations for that camera, and when those stored locations for the respective cameras all correspond to the same operating position in the work area.
21. A system for detecting the position of a torque tool as defined in any of claims 10 to 20, further comprising at least one torque tool as defined in any of claims 1 to 9.

-15 -
22. A method of detecting the position of a manually positionable torque tool, comprising: using a camera to view a work area, the camera being sensitive to light emitted by a light source on a head of the torque tool; storing a set of locations within the field of view of the camera, the set of locations corresponding to a set of operating positions in the work area; applying the torque tool to a fastening in the work area; and generating a coincidence signal if the sensed location of the light source on the torque tool coincides with one of the stored locations.
23. A method according to claim 22, wherein the set of locations is stored in a defined sequence, further comprising: applying the torque tool to a sequence of fastenings in the work area; and generating the coincidence signal each time the sensed location of the light source on the torque tool coincides with the next in sequence of the stored locations.
24. A method according to claim 22 or claim 23, further comprising: storing a torque tool characteristic in association with each of the locations; causing the intensity of light emitted by the light source on the torque tool to vary in a pattern that indicates a characteristic of the torque tool; using the camera to detect the pattern in the intensity of light; and generating the coincidence signal only when the detected pattern of light intensity represents the torque tool characteristic that is associated with the sensed location.
25. A method according to any of claims 22 to 24, further comprising disabling operation of the torque tool until the coincidence signal is generated.
26. A system for detecting the position of a manually positionable torque tool substantially as described herein with reference to the drawings.